Pasar al contenido principal

COMPUTACION DE ALTAS PRESTACIONES VII

TIN2015-65316-P

•
Nombre agencia financiadora Ministerio de Economía y Competitividad
Acrónimo agencia financiadora MINECO
Programa Programa Estatal de Fomento de la Investigación Científica y Técnica de Excelencia
Subprograma Subprograma Estatal de Generación de Conocimiento
Convocatoria Proyectos de I+D (Excelencia / Generación de Conocimiento)
Año convocatoria 2015
Plan Estatal 2013-2016 Plan Estatal de Investigación Científica y Técnica y de Innovación
Unidad de gestión Secretaría de Estado de Educación, Formación Profesional y Universidades
Centro beneficiario UNIVERSITAT POLITECNICA DE CATALUNYA
Identificador persistente http://dx.doi.org/10.13039/501100003329

Resultados relacionados

Resultados totales (Incluyendo duplicados): 516
Encontrada(s) 6 página(s)

DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Contribución de congreso (ConferenceOutput)

How2Sign: A large-scale multimodal dataset for continuous American sign language

  • Duarte, Amanda
  • Palaskar, Shruti
  • Ventura, Lucas
  • Ghadiyaram, Deepti
  • DeHaan, Kenneth
  • Metze, Florian
  • Torres, Jordi
  • Giró i Nieto, Xavier
Trabajo presentado en la IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), celebrada de forma virtual del 19 al 25 de junio de 2021, One of the factors that have hindered progress in the areas of sign language recognition, translation, and production is the absence of large annotated datasets. Towards this end, we introduce How2Sign, a multimodal and multiview continuous American Sign Language (ASL) dataset, consisting of a parallel corpus of more than 80 hours of sign language videos and a set of corresponding modalities including speech, English transcripts, and depth. A three-hour subset was further recorded in the Panoptic studio enabling detailed 3D pose estimation. To evaluate the potential of How2Sign for real-world impact, we conduct a study with ASL signers and show that synthesized videos using our dataset can indeed be understood. The study further gives insights on challenges that computer vision should address in order to make progress in this field. Dataset website: http://how2sign.github.io/, This work received funding from Facebook through gifts to CMU and UPC; through projects TEC2016-75976-R, TIN2015- 65316-P, SEV-2015-0493 and PID2019-107255GB-C22 of the Spanish Government and 2017-SGR-1414 of Generalitat de Catalunya. This work used XSEDE’s “Bridges” system at the Pittsburgh Supercomputing Center (NSF award ACI1445606). Amanda Duarte has received support from la Caixa Foundation (ID 100010434) under the fellowship code LCF/BQ/IN18/11660029. Shruti Palaskar was supported by the Facebook Fellowship program.
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

GPU4S: Embedded GPUs in Space

  • Kosmidis, Leonidas
  • Lachaize, Jérôme
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Notebaert, Olivier
  • Cazorla Almeida, Francisco Javier
  • Steenari, David
Following the same trend of automotive and avionics, the space domain is witnessing an increase in the on-board computing performance demands. This raise in performance needs comes from both control and payload parts of the spacecraft and calls for advanced electronics able to provide high computational power under the constraints of the harsh space environment. On the non-technical side, for strategic reasons it is mandatory to get European independence on the used computing technology. In this project, which is still in its early phases, we study the applicability of embedded GPUs in space, which have shown a dramatic improvement of their performance per-watt ratio coming from their proliferation in consumer markets based on competitive European technology. To that end, we perform an analysis of the existing space application domains to identify which software domains can benefit from their use. Moreover, we survey the embedded GPU domain in order to assess whether embedded GPUs can provide the required computational power and identify the challenges which need to be addressed for their adoption in space. In this paper, we describe the steps to be followed in the project, as well as the results of our preliminary analyses in the first months of the project., This work has received funding from the the European Space Agency (ESA) under the GPU4S (GPU for Space) Project, answer to the ESA ITT AO/1-9010/17/NL/AF tenderwith title ”Low Power GPU Solutions For High Performance On-Board Data Processing” and from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No. 772773). This work has also been partially supported by the Spanish Ministry of Economy and Competitiveness (MINECO) under grant TIN2015-65316-P and the HiPEAC Network of Excellence. MINECO partially supported Leonidas Kosmidis under Juan de la Cierva Formación postdoctoral fellowship (FJCI-2017-34095) and Jaume Abella under Ramon y Cajal postdoctoral fellowship (RYC-2013-14717)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Enabling a reliable STT-MRAM main memory simulation

  • Asifuzzaman, Kazi
  • Sánchez-Verdejo, Rommel
  • Radojković, Petar
STT-MRAM is a promising new memory technology with very desirable set of properties such as non-volatility, byte-addressability and high endurance. It has the potential to become the universal memory that could be incorporated to all levels of memory hierarchy. Although STT-MRAM technology got significant attention of various major memory manufacturers, to this day, academic research of STT-MRAM main memory remains marginal. This is mainly due to the unavailability of publicly available detailed timing parameters which are required to perform a cycle accurate main memory simulation. Our study presents a detailed analysis of STT-MRAM main memory timing and propose an approach to perform a reliable system level simulation of the memory technology. We seamlessly incorporate STT-MRAM timing parameters into DRAMSim2 memory simulator and use it as a part of the simulation infrastructure of the high-performance computing (HPC) systems. Our results suggests that, STT-MRAM main memory would provide performance comparable to DRAM, while opening up various opportunities for HPC system improvements. Most importantly, our study enables researchers to conduct reliable system level research on STT-MRAM main memory, and to explore the opportunities that this technology has to offer., This work was supported by BSC, Spanish Government through Programa Severo Ochoa (SEV-2015-0493), by the Spanish Ministry of Science and Technology through TIN2015-65316-P project and by the Generalitat de Catalunya (contracts 2014-SGR-1051 and 2014-SGR-1272). This work has also received funding from the European Union's Horizon 2020 research and innovation programme under ExaNoDe project (grant agreement No 671578). The authors wish to thank Terry Hulett, Duncan Bennett and Ben Cooke from Everspin Technologies Inc., for their technical support., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Advanced synchronization techniques for task-based runtime systems

  • Álvarez Robert, David
  • Sala Penadés, Kevin|||0000-0001-8233-1185
  • Maroñas Bravo, Marcos
  • Roca Nonell, Aleix|||0000-0002-6715-3605
  • Beltran Querol, Vicenç
Task-based programming models like OmpSs-2 and OpenMP provide a flexible data-flow execution model to exploit dynamic, irregular and nested parallelism. Providing an efficient implementation that scales well with small granularity tasks remains a challenge, and bottlenecks can manifest in several runtime components. In this paper, we analyze the limiting factors in the scalability of a task-based runtime system and propose individual solutions for each of the challenges, including a wait-free dependency system and a novel scalable scheduler design based on delegation. We evaluate how the optimizations impact the overall performance of the runtime, both individually and in combination. We also compare the resulting runtime against state of the art OpenMP implementations, showing equivalent or better performance, especially for fine-grained tasks., This project is supported by the European Union’s Horizon 2020 Research and Innovation programme under grant agreement No.s 754304 (DEEP-EST), by the Spanish Ministry of Science and Innovation (contract PID2019-107255GB and TIN2015-65316P) and by the Generalitat de Catalunya (2017-SGR-1414)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Runtime-assisted cache coherence deactivation in task parallel programs

  • Caheny, Paul
  • Álvarez Martí, Lluc|||0000-0003-0506-8867
  • Valero Cortés, Mateo|||0000-0003-2917-2482
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Casas, Marc|||0000-0003-4564-2093
With increasing core counts, the scalability of directory-based cache coherence has become a challenging problem. To reduce the area and power needs of the directory, recent proposals reduce its size by classifying data as private or shared, and disable coherence for private data. However, existing classification methods suffer from inaccuracies and require complex hardware support with limited scalability. This paper proposes a hardware/software co-designed approach: the runtime system identifies data that is guaranteed by the programming model semantics to not require coherence and notifies the microarchitecture. The microarchitecture deactivates coherence for this private data and powers off unused directory capacity. Our proposal reduces directory accesses to just 26% of the baseline system, and supports a 64x smaller directory with only 2.8% performance degradation. By dynamically calibrating the directory size our proposal saves 86% of dynamic energy consumption in the directory without harming performance., This work has been supported by the RoMoL ERC Advanced Grant (GA 321253), by the European HiPEAC Network of Excellence, by the Spanish Ministry of Economy and Competitiveness (contract TIN2015-65316-P), by the Generalitat de Catalunya (contracts 2014-SGR-1051 and 2014-SGR-1272) and by the European Unions Horizon 2020 research and innovation programme (grant agreements 671697 and 779877). M. Moreto has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramon y Cajal fellowship number RYC-2016-21104., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

HPC benchmarking: scaling right and looking beyond the average

  • Radulović, Milan
  • Asifuzzaman, Kazi
  • Carpenter, Paul Matthew
  • Radojković, Petar
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
The final publication is available at Springer via http://dx.doi.org/10.1007/978-3-319-96983-1_10, Designing a balanced HPC system requires an understanding of the dominant performance bottlenecks. There is as yet no well established methodology for a unified evaluation of HPC systems and workloads that quantifies the main performance bottlenecks. In this paper, we execute seven production HPC applications on a production HPC platform, and analyse the key performance bottlenecks: FLOPS performance and memory bandwidth congestion, and the implications on scaling out. We show that the results depend significantly on the number of execution processes and granularity of measurements. We therefore advocate for guidance in the application suites, on selecting the representative scale of the experiments. Also, we propose that the FLOPS performance and memory bandwidth should be represented in terms of the proportions of time with low, moderate and severe utilization. We show that this gives much more precise and actionable evidence than the average., This work was supported by the Spanish Ministry of Science and Technology (project TIN2015-65316-P), Generalitat de Catalunya (contracts 2014-SGR-1051 and 2014-SGR-1272), Severo Ochoa Programme (SEV-2015-0493) of the Spanish Government; and the European Union's Horizon 2020 research and innovation programme under ExaNoDe project (grant agreement No 671578)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Efficient Scalable Computing through Flexible Applications and Adaptive Workloads

  • Iserte, Sergio
  • Mayo, Rafael
  • Quintana Ortí, Enrique Salvador
  • Beltran Querol, Vicenç
  • Peña, Antonio J.
In this paper we introduce a methodology for dynamic job reconfiguration driven by the programming model runtime in collaboration with the global resource manager. We improve the system throughput by exploiting malleability techniques (in terms of number of MPI ranks) through the reallocation of resources assigned to a job during its execution. In our proposal, the OmpSs runtime reconfigures the number of MPI ranks during the execution of an application in cooperation with the Slurm workload manager. In addition, we take advantage of OmpSs offload semantics to allow application developers deal with data redistribution. By combining these elements a job is able to expand itself in order to exploit idle nodes or be shrunk if other queued jobs could be initiated. This novel approach adapts the system workload in order to increase the throughput as well as make a smarter use of the underlying resources. Our experiments demonstrate that this approach can reduce the total execution time of a practical workload by more than 40% while reducing the amount of resources by 30%., This work is supported by the Project TIN2014-53495-R and TIN2015-65316-P from MINECO and FEDER. Antonio J. Peña is cofinanced by MINECO under Juan de la Cierva fellowship number IJCI-2015-23266. Special thanks to José I. Aliaga for the conjugate gradient code., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

An optimized predication execution for SIMD extensions

  • Barredo Ferreira, Adrián
  • Cebrián González, Juan Manuel
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Casas, Marc|||0000-0003-4564-2093
  • Valero Cortés, Mateo|||0000-0003-2917-2482
Vector processing is a widely used technique to improve performance and energy efficiency in modern processors. Most of them rely on predication to support divergence control. However, performance and energy consumption in predicated instructions are usually independent on the number of true values in a mask. This means that the efficiency of the system becomes sub-optimal as vector length increases. In this work we propose the Optimized Predication Execution (OPE) technique. OPE delays the execution of sparse masked vector instructions sharing the same PC, extracts their active elements and creates a new dense instruction with a higher mask density. After executing such dense instruction, results are restored to the original sparse instructions. Our approach improves performance by up to 25% and reduces dynamic energy consumption by up to 43% on real applications with predication., This work has been partially supported by the RoMoL ERC Advanced Grant (GA 321253), the European HiPEAC Network of Excellence and the Spanish Government (contract TIN2015-65316-P). A. Barredo has been supported by the Spanish Government under Formación del Personal Investigador fellowship number BES-2017-080635. M. Moretó has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramon y Cajal fellowship number RYC-2016-21104., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Modelling probabilistic cache representativeness in the presence of arbitrary access patterns

  • Milutinovic, Suzana|||0000-0003-2429-0949
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Cazorla Almeida, Francisco Javier
Measurement-Based Probabilistic Timing Analysis (MBPTA) is a promising powerful industry-friendly method to derive worst-case execution time (WCET) estimates as needed for critical real-time embedded systems. MBPTA performs several (R) runs of the program on the target platform collecting the execution times in each run. MBPTA builds a probabilistic representativeness argument on whether those events with high impact on execution time, such as cache misses, arise on the runs made at analysis time so that their impact on execution time is captured. So far only events occurring in cache memories have been shown to challenge providing such representativeness argument. In this context, this paper introduces a representativeness validation method (RVS) to assess the probabilistic representativeness of MBPTA’s execution time observations in terms of cache behaviour. RVS resorts to cache simulation to predict worst-case miss scenarios that can appear during the deployment phase. RVS also constructs a probabilistic Worst-Case Miss Count curve based on the miss-counts captured in the R runs. If that curve upperbounds the impact of the predicted cache worst-case scenarios, R is deemed as a sufficient number of runs for which pWCET estimates can be reliably derived. Otherwise, the user is requested to perform more runs until all cache scenarios of interest are captured., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Design space exploration of next-generation HPC machines

  • Gómez Crespo, Constantino
  • Martínez Palau, Francesc
  • Armejach Sanosa, Adrià|||0000-0003-2869-668X
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Mantovani, Filippo|||0000-0003-3559-4825
  • Casas, Marc|||0000-0003-4564-2093
The landscape of High Performance Computing (HPC) system architectures keeps expanding with new technologies and increased complexity. With the goal of improving the efficiency of next-generation large HPC systems, designers require tools for analyzing and predicting the impact of new architectural features on the performance of complex scientific applications at scale. We simulate five hybrid (MPI+OpenMP) applications over 864 architectural proposals based on stateof-the-art and emerging HPC technologies, relevant both in industry and research. This paper significantly extends our previous work with MUltiscale Simulation Approach (MUSA) enabling accurate performance and power estimations of large-scale HPC systems. We reveal that several applications present critical scalability issues mostly due to the software parallelization approach. Looking at speedup and energy consumption exploring the design space (i.e., changing memory bandwidth, number of cores, and type of cores), we provide evidence-based architectural recommendations that will serve as hardware and software codesign guidelines., This work is partially supported by the Spanish Gov-ernment through Programa Severo Ochoa (SEV-2015-0493),by the Spanish Ministry of Science and Technology project(TIN2015-65316-P), by the Generalitat de Catalunya (2017-SGR-1414 and 2017-SGR-1328), and by the European Com-munity’s Horizon 2020 Framework Programme under theMont-Blanc project, grant agreement n. (671697 and 779877).A. Armejach has been partially supported by the SpanishMinistry of Economy, Industry and Competitiveness underJuan de la Cierva postdoctoral fellowship number FJCI-2015-24753 M. Moret ́o has been partially supported by the SpanishMinistry of Economy, Industry and Competitiveness underRamon y Cajal fellowship number RYC-2016-21104. M.Casas has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramon y Cajal fellowship number RYC-2017-23269., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Characterizing performance and energy-efficiency of the RAMCloud storage system

  • Taleb, Yacine
  • Ibrahim, Shadi
  • Antoniu, Gabriel
  • Cortés, Toni|||0000-0002-2537-8937
Most large popular web applications, like Facebook and Twitter, have been relying on large amounts of in-memory storage to cache data and offer a low response time. As the main memory capacity of clusters and clouds increases, it becomes possible to keep most of the data in the main memory. This motivates the introduction of in-memory storage systems. While prior work has focused on how to exploit the low-latency of in-memory access at scale, there is very little visibility into the energy-efficiency of in-memory storage systems. Even though it is known that main memory is a fundamental energy bottleneck in computing systems (i.e., DRAM consumes up to 40% of a server's power). In this paper, by the means of experimental evaluation, we have studied the performance and energy-efficiency of RAMCloud - a well-known in-memory storage system. We reveal that although RAMCloud is scalable for read-only applications, it exhibits non-proportional power consumption. We also find that the current replication scheme implemented in RAMCloud limits the performance and results in high energy consumption. Surprisingly, we show that replication can also play a negative role in crash-recovery., This work has been supported by the BigStorage project, funded by the European Union under the Marie Sklodowska-Curie Actions (H2020-MSCA-ITN-2014-642963), by the Spanish Government (grant SEV2015-1305 0493 of the Severo Ochoa Program), by the Spanish Ministry of Science and Innovation (contract TIN2015-65316), by Generalitat de Catalunya (contract 2014-SGR-1051), and by the ANR KerStream project (ANR-16- CE25-0014-01). The experiments presented in this paper were carried out using the Grid5000/ALADDIN-G5K experimental testbed, an initiative from the French Ministry of Research through the ACI GRID incentive action, INRIA, CNRS and RENATER and other contributing partners (see http://www.grid5000.fr/ for details)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

RePP-C: runtime estimation of performance-power with workload consolidation in CMPs

  • Nishtala, Rajiv
  • Martorell Bofill, Xavier|||0000-0002-0417-3430
Configuration of hardware knobs in multicore environments for meeting performance-power demands constitutes a desirable feature in modern data centers. At the same time, high energy efficiency (performance per watt) requires optimal thread-to-core assignment. In this paper, we present the runtime estimator (RePP-C) for performance-power, characterized by processor frequency states (P-states), a wide range of sleep intervals (Cl-states) and workload consolidation. We also present a schema for frequency and contention-aware thread-to-core assignment (FACTS) which considers various thread demands. The proposed solution (RePP-C) selects a given hardware configuration for each active core to ensure that the performance-power demands are satisfied while using the scheduling schema (FACTS) for mapping threads-to-cores. Our results show that FACTS improves over other state-of-the-art schedulers like Distributed Intensity Online (DIO) and native Linux scheduler by 8.25% and 37.56% in performance, with simultaneous improvement in energy efficiency by 6.2% and 14.17%, respectively. Moreover, we prove the usability of RePP-C by predicting performance and power for 7 different types of workloads and 10 different QoS targets. The results show an average error of 7.55% and 8.96% (with 95% confidence interval) when predicting energy and performance respectively., This work has been partially supported by the European Union FP7 program through the Mont-Blanc-2 project (FP7-ICT-610402), by the Ministerio de Economia y Competitividad under contract Computacion de Altas Prestaciones VII (TIN2015-65316-P), and the Departament d’Innovacio, Universitats i Empresa de la Generalitat de Catalunya, under project MPEXPAR: Models de Programacio i Entorns d’Execucio Paral.lels (2014-SGR-1051)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Performance evaluation of data-centric workloads in serverless environments

  • Nestorov, Anna Maria
  • Polo Bardés, Jorda
  • Misale, Claudia
  • Carrera Pérez, David|||0000-0003-4898-3424
  • Youssef, Alaa
Serverless computing is a cloud-based execution paradigm that allows provisioning resources on-demand, freeing developers from infrastructure management and operational concerns. It typically involves deploying workloads as stateless functions that take no resources when not in use, and is meant to scale transparently. To make serverless effective, providers impose limits on a per-function level, such as maximum duration, fixed amount of memory, and no persistent local storage. These constraints make it challenging for data-intensive workloads to take advantage of serverless because they lead to sharing significant amounts of data through remote storage. In this paper, we build a performance model for serverless workloads that considers how data is shared between functions, including the amount of data and the underlying technology that is being used. The model's accuracy is assessed by running a real workload in a cluster using Knative, a state-of-the-art serverless environment, showing a relative error of 5.52%. With the proposed model, we evaluate the performance of data-intensive workloads in serverless, analyzing parallelism, scalability, resource requirements, and scheduling policies. We also explore possible solutions for the data-sharing problem, like using local memory and storage. Our results show that the performance of data-intensive workloads in serverless can be up to 4.32= faster depending on how these are deployed., This work was partially supported by the Ministry of Economy of Spain under contract TIN2015-65316-P, the Ministry of Science under contract PID2019-107255GB-C21/AEI/10.13039/501100011033, and the Generalitat de Catalunya under contract 2014SGR1051., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Assessing the Adherence of an Industrial Autonomous Driving Framework to ISO 26262 Software Guidelines

  • Tabani, Hamid
  • Kosmidis, Leonidas
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Cazorla, Francisco J.
  • Bernat, Guillem
The complexity and size of Autonomous Driving (AD) software are comparably higher than that of software implementing other (standard) functionalities in the car. To make things worse, a big fraction of AD software is not specifically designed for the automotive (or any other critical) domain, but the mainstream market. This brings uncertainty on to which extent AD software adheres to guidelines in safety standards. In this paper, we present our experience in applying ISO 26262 -- the applicable functional safety standard for road vehicles -- software safety guidelines to industrial AD software, in particular, Apollo, a heterogeneous Autonomous Driving framework used extensively in industry. We provide quantitative and qualitative metrics of compliance for many ISO 26262 recommendations on software design, implementation, and testing., This work has received funding from the European Research Coun-cil (ERC) under the European Union’s Horizon 2020 research andinnovation programme (grant agreement No. 772773). This workhas also been partially supported by the Spanish Ministry of Econ-omy and Competitiveness (MINECO) under grant TIN2015-65316-Pand the HiPEAC Network of Excellence. MINECO partially sup-ported Jaume Abella under Ramon y Cajal postdoctoral fellowship(RYC-2013-14717), and Leonidas Kosmidis under Juan de la Cierva-Formación postdoctoral fellowship (FJCI-2017-34095)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Go with the flow: Recurrent networks for wind time series multi-step forecasting

  • Manero Font, Jaume
  • Béjar Alonso, Javier|||0000-0001-5281-3888
  • Cortés García, Claudio Ulises|||0000-0003-0192-3096
One of the ways of reducing the effects of Climate Change is to rely on renewable energy sources. Their intermittent nature makes necessary to obtain a mid-long term accurate forecasting. Wind Energy prediction is based on the ability to forecast wind speed. This has been a problem approached using different methods based on the statistical properties of the wind time series. Wind Time series are non-linear and non-stationary, making their forecasting very challenging. Deep neural networks have shown their success recently for problems involving sequences with non-linear behavior. In this work, we perform experiments comparing the capability of different neural network architectures for multi-step forecasting obtaining a 12 hours ahead prediction using data from the National Renewable Energy Laboratory's WIND dataset, Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Continuous-action reinforcement learning for memory allocation in virtualized servers

  • Garrido Platero, Luis Ángel
  • Nishtala, Rajiv
  • Carpenter, Paul Matthew
In a virtualized computing server (node) with multiple Virtual Machines (VMs), it is necessary to dynamically allocate memory among the VMs. In many cases, this is done only considering the memory demand of each VM without having a node-wide view. There are many solutions for the dynamic memory allocation problem, some of which use machine learning in some form. This paper introduces CAVMem (Continuous-Action Algorithm for Virtualized Memory Management), a proof-of-concept mechanism for a decentralized dynamic memory allocation solution in virtualized nodes that applies a continuous-action reinforcement learning (RL) algorithm called Deep Deterministic Policy Gradient (DDPG). CAVMem with DDPG is compared with other RL algorithms such as Q-Learning (QL) and Deep Q-Learning (DQL) in an environment that models a virtualized node. In order to obtain linear scaling and be able to dynamically add and remove VMs, CAVMem has one agent per VM connected via a lightweight coordination mechanism. The agents learn how much memory to bid for or return, in a given state, so that each VM obtains a fair level of performance subject to the available memory resources. Our results show that CAVMem with DDPG performs better than QL and a static allocation case, but it is competitive with DQL. However, CAVMem incurs significant less training overheads than DQL, making the continuous-action approach a more cost-effective solution., This research is part of a project that has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 754337 (EuroEXA) and the European Union’s 7th Framework Programme under grant agreement number 610456 (Euroserver). It also received funding from the Spanish Ministry of Science and Technology (project TIN2015-65316-P), Generalitat de Catalunya (contract 2014-SGR-1272), and the Severo Ochoa Programme (SEV-2015-0493) of the Spanish Government., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Runtime-guided mitigation of manufacturing variability in power-constrained multi-socket NUMA nodes

  • Chasapis, Dimitrios
  • Casas, Marc|||0000-0003-4564-2093
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Schulz, Martin
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
  • Labarta Mancho, Jesús José|||0000-0002-7489-4727
  • Valero Cortés, Mateo|||0000-0003-2917-2482
This work has been supported by the Spanish Government (Severo Ochoa grants SEV2015-0493, SEV-2011-00067), by the Spanish Ministry of Science and Innovation (contracts TIN2015-65316-P), by Generalitat de Catalunya (contracts 2014-SGR-1051 and 2014-SGR-1272), by the RoMoL ERC Advanced Grant (GA 321253) and the European HiPEAC Network of Excellence. M. Moretó has been partially supported by the Ministry of Economy and Competitiveness under Juan de la Cierva postdoctoral fellowship number JCI-2012-15047. M. Casas is supported by the Secretary for Universities and Research of the Ministry of Economy and Knowledge of the Government of Catalonia and the Cofund programme of the Marie Curie Actions of the 7th R&D Framework Programme of the European Union (Contract 2013 BP B 00243). This work was also partially performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344 (LLNL-CONF-689878). Finally, the authors are grateful to the reviewers for their valuable comments, to the RoMoL team, to Xavier Teruel and Kallia Chronaki from the Programming Models group of BSC and the Computation Department of LLNL for their technical support and useful feedback., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

TaskGenX: A Hardware-Software Proposal for Accelerating Task Parallelism

  • Chronaki, Kallia
  • Casas, Marc|||0000-0003-4564-2093
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Bosch Pons, Jaume|||0000-0002-4040-3416
  • Badia Sala, Rosa Maria|||0000-0003-2941-5499
As chip multi-processors (CMPs) are becoming more and more complex, software solutions such as parallel programming models are attracting a lot of attention. Task-based parallel programming models offer an appealing approach to utilize complex CMPs. However, the increasing number of cores on modern CMPs is pushing research towards the use of fine grained parallelism. Task-based programming models need to be able to handle such workloads and offer performance and scalability. Using specialized hardware for boosting performance of task-based programming models is a common practice in the research community. Our paper makes the observation that task creation becomes a bottleneck when we execute fine grained parallel applications with many task-based programming models. As the number of cores increases the time spent generating the tasks of the application is becoming more critical to the entire execution. To overcome this issue, we propose TaskGenX. TaskGenX offers a solution for minimizing task creation overheads and relies both on the runtime system and a dedicated hardware. On the runtime system side, TaskGenX decouples the task creation from the other runtime activities. It then transfers this part of the runtime to a specialized hardware. We draw the requirements for this hardware in order to boost execution of highly parallel applications. From our evaluation using 11 parallel workloads on both symmetric and asymmetric multicore systems, we obtain performance improvements up to 15×, averaging to 3.1× over the baseline., This work has been supported by the RoMoL ERC Advanced Grant (GA 321253), by the European HiPEAC Network of Excellence, by the Spanish Ministry of Science and Innovation (contracts TIN2015-65316-P), by the Generalitat de Catalunya (contracts 2014-SGR-1051 and 2014-SGR-1272), and by the European Union’s Horizon 2020 research and innovation programme under grant agreement No. 671697 and No. 779877. M. Moretó has been partially supported by the Ministry of Economy and Competitiveness under Ramon y Cajal fellowship number RYC-2016-21104. Finally, the authors would like to thank Thomas Grass for his valuable help with the simulator., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

RPR: a random replacement policy with limited pathological replacements

  • Benedicte Illescas, Pedro|||0000-0003-1670-7783
  • Hernández Luz, Carles
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Cazorla Almeida, Francisco Javier
Measurement-Based Probabilistic Timing Analysis (MBPTA) has consolidated as a technique to estimate probabilistic Worst-Case Execution Times (WCET) for critical software running on processors with high-performance hardware such as multilevel caches. Conventional random replacement (CRR) is the most suitable replacement policy for MBPTA due to its probabilistic nature: replacement choices are random and independent. CRR makes pathological replacement patterns probabilistic rather than systematic, though they can still occur. This paper proposes a new replacement policy, RPR, that keeps MBPTA compatibility and prevents CRR's pathological replacements in which addresses mapped to different cache lines randomly evict each other despite some lines in the same cache set are available. In particular, RPR maintains a higher degree of temporal locality than CRR. Our evaluation on a performance simulator (validated against a real industrial prototype) using the Mälardalen benchmarks and a railway case study shows that RPR delivers both high average performance (within 1% of LRU's performance) and tight WCET estimates 16% and 24% (for the case study and Mälardalen respectively) lower than those of CRR., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

CATA: Criticality aware task acceleration for multicore processors

  • Castillo, Emilio
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Casas, Marc|||0000-0003-4564-2093
  • Álvarez Martí, Lluc|||0000-0003-0506-8867
  • Vallejo, Enrique
  • Chronaki, Kallia
  • Badia Sala, Rosa Maria|||0000-0003-2941-5499
  • Bosque Orero, José Luis
  • Beivide Palacio, Julio Ramón
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
  • Labarta Mancho, Jesús José|||0000-0002-7489-4727
  • Valero Cortés, Mateo|||0000-0003-2917-2482
Managing criticality in task-based programming models opens a wide range of performance and power optimization opportunities in future manycore systems. Criticality aware task schedulers can benefit from these opportunities by scheduling tasks to the most appropriate cores. However, these schedulers may suffer from priority inversion and static binding problems that limit their expected improvements. Based on the observation that task criticality information can be exploited to drive hardware reconfigurations, we propose a Criticality Aware Task Acceleration (CATA) mechanism that dynamically adapts the computational power of a task depending on its criticality. As a result, CATA achieves significant improvements over a baseline static scheduler, reaching average improvements up to 18.4% in execution time and 30.1% in Energy-Delay Product (EDP) on a simulated 32-core system. The cost of reconfiguring hardware by means of a software-only solution rises with the number of cores due to lock contention and reconfiguration overhead. Therefore, novel architectural support is proposed to eliminate these overheads on future manycore systems. This architectural support minimally extends hardware structures already present in current processors, which allows further improvements in performance with negligible overhead. As a consequence, average improvements of up to 20.4% in execution time and 34.0% in EDP are obtained, outperforming state-of-the-art acceleration proposals not aware of task criticality., This work has been supported by the Spanish Government (grant SEV2015-0493, SEV-2011-00067 of the Severo Ochoa Program), by the Spanish Ministry of Science and Innovation (contracts TIN2015-65316, TIN2012-34557, TIN2013-46957-C2-2-P), by Generalitat de Catalunya (contracts 2014-SGR- 1051 and 2014-SGR-1272), by the RoMoL ERC Advanced Grant (GA 321253) and the European HiPEAC Network of Excellence. The Mont-Blanc project receives funding from the EU’s Seventh Framework Programme (FP7/2007-2013) under grant agreement no 610402 and from the EU’s H2020 Framework Programme (H2020/2014-2020) under grant agreement no 671697. M. Moret´o has been partially supported by the Ministry of Economy and Competitiveness under Juan de la Cierva postdoctoral fellowship number JCI-2012-15047. M. Casas is supported by the Secretary for Universities and Research of the Ministry of Economy and Knowledge of the Government of Catalonia and the Cofund programme of the Marie Curie Actions of the 7th R&D Framework Programme of the European Union (Contract 2013 BP B 00243). E. Castillo has been partially supported by the Spanish Ministry of Education, Culture and Sports under grant FPU2012/2254., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Towards a goal-oriented agent-based simulation framework for high-performance computing

  • Gnatyshak, Dmitry|||0000-0001-6779-6283
  • Oliva Felipe, Luis Javier|||0000-0001-6555-1549
  • Álvarez Napagao, Sergio|||0000-0001-9946-9703
  • Padget, Julián
  • Vázquez Salceda, Javier|||0000-0003-1732-9446
  • Garcia Gasulla, Dario|||0000-0001-6732-5641
  • Cortés García, Claudio Ulises|||0000-0003-0192-3096
Currently, agent-based simulation frameworks force the user to choose between simulations involving a large number of agents (at the expense of limited agent reasoning capability) or simulations including agents with increased reasoning capabilities (at the expense of a limited number of agents per simulation). This paper describes a first attempt at putting goal-oriented agents into large agentbased (micro-)simulations. We discuss a model for goal-oriented agents in HighPerformance Computing (HPC) and then briefly discuss its implementation in PyCOMPSs (a library that eases the parallelisation of tasks) to build such a platform that benefits from a large number of agents with the capacity to execute complex cognitive agents., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Design and implementation of a fair credit-based bandwidth sharing scheme for buses

  • Slijepcevic, Mladen
  • Hernandez, Carles
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Cazorla, Francisco J.
Fair arbitration in the access to hardware shared resources is fundamental to obtain low worst-case execution time (WCET) estimates in the context of critical real-time systems, for which performance guarantees are essential. Several hardware mechanisms exist for managing arbitration in those resources (buses, memory controllers, etc.). They typically attain fairness in terms of the number of slots each contender (e.g., core) gets granted access to the shared resource. However, those policies may lead to unfair bandwidth allocations for workloads with contenders issuing short requests and contenders issuing long requests. We propose a Credit-Based Arbitration (CBA) mechanism that achieves fairness in the cycles each core is granted access to the resource rather than in the number of granted slots. Furthermore, we implement CBA as part of a LEON3 4-core processor for the Space domain in an FPGA proving the feasibility and good performance characteristics of the design by comparing it against other arbitration schemes., The research leading to these results has received funding from the European Community’s Seventh Framework Programme [FP7/2007-2013] under the PROXIMA Project (www.proxima-project.eu), grant agreement no 611085. This work has also been partially supported by the Spanish Ministry of Science and Innovation under grant TIN2015-65316-P and the HiPEAC Network of Excellence. Mladen Slijepcevic is funded by the Obra Social Fundaci´on la Caixa under grant Doctorado “la Caixa” - Severo Ochoa. Carles Hernández is jointly funded by the Spanish Ministry of Economy and Competitiveness (MINECO) and FEDER funds through grant TIN2014-60404-JIN. Jaume Abella has been partially supported by the MINECO under Ramon y Cajal postdoctoral fellowship number RYC-2013-14717., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Performance analysis of a hardware accelerator of dependence management for taskbased dataflow programming models

  • Tan, Xubin
  • Bosch Pons, Jaume|||0000-0002-4040-3416
  • Jiménez González, Daniel|||0000-0001-6064-7883
  • Álvarez Martínez, Carlos|||0000-0003-0536-5183
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
  • Valero Cortés, Mateo|||0000-0003-2917-2482
Along with the popularity of multicore and manycore, task-based dataflow programming models obtain great attention for being able to extract high parallelism from applications without exposing the complexity to programmers. One of these pioneers is the OpenMP Superscalar (OmpSs). By implementing dynamic task dependence analysis, dataflow scheduling and out-of-order execution in runtime, OmpSs achieves high performance using coarse and medium granularity tasks. In theory, for the same application, the more parallel tasks can be exposed, the higher possible speedup can be achieved. Yet this factor is limited by task granularity, up to a point where the runtime overhead outweighs the performance increase and slows down the application. To overcome this handicap, Picos was proposed to support task-based dataflow programming models like OmpSs as a fast hardware accelerator for fine-grained task and dependence management, and a simulator was developed to perform design space exploration. This paper presents the very first functional hardware prototype inspired by Picos. An embedded system based on a Zynq 7000 All-Programmable SoC is developed to study its capabilities and possible bottlenecks. Initial scalability and hardware consumption studies of different Picos designs are performed to find the one with the highest performance and lowest hardware cost. A further thorough performance study is employed on both the prototype with the most balanced configuration and the OmpSs software-only alternative. Results show that our OmpSs runtime hardware support significantly outperforms the software-only implementation currently available in the runtime system for finegrained tasks., This work is supported by the Spanish Government through Programa Severo Ochoa (SEV-2015-0493), by the Spanish Ministry of Science and Technology through TIN2015-65316-P project, by the Generalitat de Catalunya (contracts 2014-SGR-1051 and 2014-SGR-1272) and by the European Research Council RoMoL Grant Agreement number 321253. We also thank the Xilinx University Program for its hardware and software donations., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Modeling RTL Fault Models Behavior to Increase the Confidence on TSIM-based Fault Injection

  • Espinosa, Jaime
  • Hernandez, Carles
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
Future high-performance safety-relevant applications require microcontrollers delivering higher performance than the existing certified ones. However, means for assessing their dependability are needed so that they can be certified against safety critical certification standars (e.g ISO26262). Dependability assessment analyses performed at high level of abstraction inject single faults to investigate the effects these have in the system. In this work we show that single faults do not comprise the whole picture, due to fault multiplicities and reactivations. Later we prove that, by injecting complex fault models that consider multiplicities and reactivations in higher levels of abstraction, results are substantially different, thus indicating that a change in the methodology is needed., The research leading to these results has received funding from the Ministry of Science and Technology of Spain under contract TIN2015-65316-P and the HiPEAC Network of Excellence. Carles Hern´andez is jointly funded by the Spanish Ministry of Economy and Competitiveness (MINECO) and FEDER funds through grant TIN2014-60404-JIN. Jaume Abella has been partially supported by the MINECO under Ramon y Cajal postdoctoral fellowship number RYC-2013-14717.
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Multiple target task sharing support for the OpenMP accelerator model

  • Ozen, Guray|||0000-0001-8237-0141
  • Mateo, Sergi
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
  • Labarta Mancho, Jesús José|||0000-0002-7489-4727
  • Beyer, James B.
The use of GPU accelerators is becoming common in HPC platforms due to the their effective performance and energy efficiency. In addition, new generations of multicore processors are being designed with wider vector units and/or larger hardware thread counts, also contributing to the peak performance of the whole system. Although current directive–based paradigms, such as OpenMP or OpenACC, support both accelerators and multicore-based hosts, they do not provide an effective and efficient way to concurrently use them, usually resulting in accelerated programs in which the potential computational performance of the host is not exploited. In this paper we propose an extension to the OpenMP 4.5 directive-based programming model to support the specification and execution of multiple instances of task regions on different devices (i.e. accelerators in conjunction with the vector and heavily multithreaded capabilities in multicore processors). The compiler is responsible for the generation of device-specific code for each device kind, delegating to the runtime system the dynamic schedule of the tasks to the available devices. The new proposed clause conveys useful insight to guide the scheduler while keeping a clean, abstract and machine independent programmer interface. The potential of the proposal is analyzed in a prototype implementation in the OmpSs compiler and runtime infrastructure. Performance evaluation is done using three kernels (N-Body, tiled matrix multiply and Stream) on different GPU-capable systems based on ARM, Intel x86 and IBM Power8. From the evaluation we observe speed–ups in the 8–20% range compared to versions in which only the GPU is used, reaching 96 % of the additional peak performance thanks to the reduction of data transfers and the benefits introduced by the OmpSs NUMA-aware scheduler., This work is partially supported by the IBM/BSC Deep Learning Center Initiative, by the Spanish Government through Programa Severo Ochoa (SEV-2015-0493), by the Spanish Ministry of Science and Technology through TIN2015-65316-P project and by the Generalitat de Catalunya (contract 2014-SGR-1051)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

On uses of extreme value theory fit for industrial-quality WCET analysis

  • Milutinovic, Suzana|||0000-0003-2429-0949
  • Mezzetti, Enrico|||0000-0002-1886-2931
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Vardanega, Tullio
  • Cazorla, Francisco J.
Over the last few years, considerable interest has arisen in measurement-based probabilistic timing analysis. The term MBPTA has been used to indistinctly refer to a variety of different applications of Extreme Value Theory (EVT) to the timing analysis problem. The successful application of MBPTA techniques to a score of case studies has not fully dispelled the concerns that industrial stakeholders had with the quality of the computed bounds, hence ultimately with their industrial viability. Placing focus on the MBPTA methods and techniques developed in the PROARTIS and PROXIMA projects, collectively referred to as proMBPTA, we discuss the main misconceptions and pitfalls that can prevent a sound application of EVT-based WCET analysis. Using a combination of arguments and support examples, we show that proMBPTA is a rigorous process, fully amenable to sound and sustainable industrial use., This work has been partially supported by the Spanish Ministry of Economy and Competitiveness (MINECO) under grant TIN2015-65316-P and the HiPEAC Network of Excellence. Jaume Abella has been partially supported by the MINECO under Ramon y Cajal grant RYC-2013-14717. Authors also thank George Lima for his feedback on this manuscript., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Rebalancing the core front-end through HPC code analysis

  • Milic, Ugljesa
  • Carpenter, Paul Matthew
  • Rico, Alejandro
  • Ramirez, Alex
There is a need to increase performance under the same power and area envelope to achieve Exascale technology in high performance computing (HPC). The today's chip multiprocessor (CMP) design is tailored by traditional desktop and server workloads, different from parallel applications commonly run in HPC. In this work, we focus on the HPC code characteristics and processor front-end which factors around 30% of core power and area on the emerging lean-core type of processors used in HPC. Separating serial from parallel code sections inside applications, we characterize three HPC benchmark suites and compare them to a traditional set of desktop integer workloads. HPC applications have biased and mostly backward taken branches, small dynamic instruction footprints, and long basic blocks. Our findings suggest smaller branch predictors (BP) with the additional loop BP, smaller branch target buffers (BTB), and smaller L1 instruction caches (I-cache) with wider lines. Still, the aforementioned downsizing applies only to the cores meant to run parallel code. The difference between serial and parallel code sections in HPC applications points to an asymmetric CMP design, with one baseline core for sequential and many HPCtailored cores designed for parallel code. Predictions using Sniper simulator and McPAT show that an HPC-tailored lean core saves 16% of the core area and 7% of power compared to a baseline core, without performance loss. Using the area savings to add an extra core, an asymmetric CMP with one baseline and eight tailored cores has the same area budget as a symmetric CMP composed out of eight baseline cores demanding 4% more power and providing 12% shorter execution time on average., The research was supported by European Unions 7th Framework Programme [FP7/2007-2013] under project Mont-Blanc (288777), the Ministry of Economy and Competitiveness of Spain (TIN2012-34557, TIN2015-65316-P, and BES-2013-063925), Generalitat de Catalunya (2014-SGR-1051 and 2014-SGR-1272), HiPEAC-3 Network of Excellence (ICT-287759), and finally the Severo Ochoa Program (SEV-2011-00067) of the Spanish Government., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Mitigating software-instrumentation cache effects in measurement-based timing analysis

  • Díaz, Enrique
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Mezzetti, Enrico|||0000-0002-1886-2931
  • Aguirre, Irune
  • Azkarate-Askasua, Mikel
  • Vardanega, Tullio
  • Cazorla, Francisco
Measurement-based timing analysis (MBTA) is often used to determine the timing behaviour of software programs embedded in safety-aware real-time systems deployed in various industrial domains including automotive and railway. MBTA methods rely on some form of instrumenta- tion, either at hardware or software level, of the target program or fragments thereof to collect execution-time measurement data. A known drawback of software-level instrumentation is that instrumentation itself does affect the timing and functional behaviour of a program, resulting in the so-called probe effect: leaving the instrumentation code in the final executable can negatively affect average performance and could not be even admissible under stringent industrial qualification and certification standards; removing it before operation jeopardizes the results of timing analysis as the WCET estimates on the instrumented version of the program cannot be valid any more due, for example, to the timing effects incurred by different cache alignments. In this paper, we present a novel approach to mitigate the impact of instrumentation code on cache behaviour by reducing the instrumentation overhead while at the same time preserving and consolidating the results of timing analysis., The research leading to these results has received funding from the European Community’s FP7 [FP7/2007-2013] under the PROXIMA Project (http://www.proxima-project.eu), grant agreement no. 611085. This work has also been partially supported by the Spanish Ministry of Science and Innovation (grant TIN2015-65316-P) and the HiPEAC Network of Excellence. Jaume Abella has been partially supported by the Ministry of Economy and Competitiveness under Ramon y Cajal fellowship RYC-2013-14717., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Probabilistic timing analysis on time-randomized platforms for the space domain

  • Fernandez, Mikel
  • Morales, David
  • Kosmidis, Leonidas
  • Bardizbanyan, Alen
  • Broster, Ian
  • Hernandez, Carles
  • Quiñones, Eduardo
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Cazorla, Francisco
  • Machado, Paulo
  • Fossati, Luca
Timing Verification is a fundamental step in real-time embedded systems, with measurement-based timing analysis (MBTA) being the most common approach used to that end. We present a Space case study on a real platform that has been modified to support a probabilistic variant of MBTA called MBPTA. Our platform provides the properties required by MBPTA with the predicted WCET estimates with MBPTA being competitive to those with current MBTA practice while providing more solid evidence on their correctness for certification., The research leading to these results has received funding from the European Community’s FP7 [FP7/2007-2013] under the PROXIMA Project (www.proxima-project.eu), grant agreement no 611085. This work has also been partially supported by the Spanish Ministry of Science and Innovation under grant TIN2015-65316-P and the HiPEAC Network of Excellence. Jaume Abella has been partially supported by the Ministry of Economy and Competitiveness under Ramon y Cajal postdoctoral fellowship number RYC-2013-14717. Carles Hernandez is jointly funded by the Spanish Ministry of Economy and Competitiveness and FEDER funds through grant TIN2014-60404-JIN., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Runtime-guided management of stacked DRAM memories in task parallel programs

  • Álvarez Martí, Lluc|||0000-0003-0506-8867
  • Casas, Marc|||0000-0003-4564-2093
  • Labarta Mancho, Jesús José|||0000-0002-7489-4727
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
  • Valero Cortés, Mateo|||0000-0003-2917-2482
  • Moretó Planas, Miquel|||0000-0002-9848-8758
Stacked DRAM memories have become a reality in High-Performance Computing (HPC) architectures. These memories provide much higher bandwidth while consuming less power than traditional off-chip memories, but their limited memory capacity is insufficient for modern HPC systems. For this reason, both stacked DRAM and off-chip memories are expected to co-exist in HPC architectures, giving raise to different approaches for architecting the stacked DRAM in the system. This paper proposes a runtime approach to transparently manage stacked DRAM memories in task-based programming models. In this approach the runtime system is in charge of copying the data accessed by the tasks to the stacked DRAM, without any complex hardware support nor modifications to the application code. To mitigate the cost of copying data between the stacked DRAM and the off-chip memory, the proposal includes an optimization to parallelize the copies across idle or additional helper threads. In addition, the runtime system is aware of the reuse pattern of the data accessed by the tasks, and can exploit this information to avoid unworthy copies of data to the stacked DRAM. Results on the Intel Knights Landing processor show that the proposed techniques achieve an average speedup of 14% against the state-of-the-art library to manage the stacked DRAM and 29% against a stacked DRAM architected as a hardware cache., This work has been supported by the RoMoL ERC Advanced Grant (GA 321253), by the European HiPEAC Network of Excellence, by the Spanish Ministry of Economy and Competitiveness (contract TIN2015-65316-P), by the Generalitat de Catalunya (contracts 2014-SGR-1051 and 2014-SGR-1272) and by the European Union’s Horizon 2020 research and innovation programme (grant agreement 779877). M. Moreto has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramon y Cajal fellowship number RYC-2016-21104., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

How2Sign: A large-scale multimodal dataset for continuous American sign language

  • Cardoso Duarte, Amanda
  • Palaskar, Shruti
  • Ventura Ripol, Lucas
  • Ghadiyaram, Deepti
  • DeHaan, Kenneth
  • Metze, Florian
  • Torres Viñals, Jordi|||0000-0003-1963-7418
  • Giró Nieto, Xavier|||0000-0002-9935-5332
One of the factors that have hindered progress in the areas of sign language recognition, translation, and production is the absence of large annotated datasets. Towards this end, we introduce How2Sign, a multimodal and multiview continuous American Sign Language (ASL) dataset, consisting of a parallel corpus of more than 80 hours of sign language videos and a set of corresponding modalities including speech, English transcripts, and depth. A three-hour subset was further recorded in the Panoptic studio enabling detailed 3D pose estimation. To evaluate the potential of How2Sign for real-world impact, we conduct a study with ASL signers and show that synthesized videos using our dataset can indeed be understood. The study further gives insights on challenges that computer vision should address in order to make progress in this field. Dataset website: http://how2sign.github.io/, This work received funding from Facebook through gifts to CMU and UPC; through projects TEC2016-75976-R, TIN2015- 65316-P, SEV-2015-0493 and PID2019-107255GB-C22 of the Spanish Government and 2017-SGR-1414 of Generalitat de Catalunya. This work used XSEDE’s “Bridges” system at the Pittsburgh Supercomputing Center (NSF award ACI- 1445606). Amanda Duarte has received support from la Caixa Foundation (ID 100010434) under the fellowship code LCF/BQ/IN18/11660029. Shruti Palaskar was supported by the Facebook Fellowship program., Peer Reviewed, Objectius de Desenvolupament Sostenible::10 - Reducció de les Desigualtats, Objectius de Desenvolupament Sostenible::4 - Educació de Qualitat::4.5 - Per a 2030, eliminar les disparitats de gènere en l’educació i garantir l’accés en condicions d’igualtat a les persones vulnerables, incloses les persones amb discapacitat, els pobles indígenes i els nens i nenes en situacions de vulnerabilitat, a tots els nivells de l’ensenyament i la formació professional, Objectius de Desenvolupament Sostenible::10 - Reducció de les Desigualtats::10.2 - Per a 2030, potenciar i promoure la inclusió social, econòmica i política de totes les persones, independentment de l’edat, sexe, discapacitat, raça, ètnia, origen, religió, situació econòmica o altra condició, Objectius de Desenvolupament Sostenible::4 - Educació de Qualitat
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

On the benefits of tasking with OpenMP

  • Rico, Alejandro
  • Sánchez Barrera, Isaac|||0000-0003-1616-0685
  • Joao, Jose A.
  • Randall, Joshua
  • Casas, Marc|||0000-0003-4564-2093
  • Moretó Planas, Miquel|||0000-0002-9848-8758
Tasking promises a model to program parallel applications that provides intuitive semantics. In the case of tasks with dependences, it also promises better load balancing by removing global synchronizations (barriers), and potential for improved locality. Still, the adoption of tasking in production HPC codes has been slow. Despite OpenMP supporting tasks, most codes rely on worksharing-loop constructs alongside MPI primitives. This paper provides insights on the benefits of tasking over the worksharing-loop model by reporting on the experience of taskifying an adaptive mesh refinement proxy application: miniAMR. The performance evaluation shows the taskified implementation being 15–30% faster than the loop-parallel one for certain thread counts across four systems, three architectures and four compilers thanks to better load balancing and system utilization. Dynamic scheduling of loops narrows the gap but still falls short of tasking due to serial sections between loops. Locality improvements are incidental due to the lack of locality-aware scheduling. Overall, the introduction of asynchrony with tasking lives up to its promises, provided that programmers parallelize beyond individual loops and across application phases., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Freezing Time: a new approach for emulating fast storage devices using VM

  • Bona, Luis C.E.
  • Elias, Alessandro
  • Ziviani, Andre P.
  • Cortés, Toni|||0000-0002-2537-8937
  • Nou Castell, Ramon|||0000-0003-0464-1473
  • Alves, Marco A.Z.
© 2018 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes,creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works., Recently we are seeing a considerable effort from both academy and industry in proposing new technologies for storage devices. Often these devices are not readily available for evaluation and methods to allow performing their tests just from their performance parameters are an important tool for system administrators. Simulators are a traditional approach for carrying out such evaluations, however, they are more suitable for evaluating the storage device as an isolate component, mostly due to time constraints. In this paper, we propose an approach based on virtual machine technology that is capable of emulate storage devices transparently for the operating system allowing evaluation of simulating devices within a real system using any synthetic or real workload. To emulate devices in real environments it is necessary to use the currently available devices as a storage medium which creates a difficulty when the device to be emulated is faster than this storage medium. To circumvent this limitation we introduce a new technique called Freezing Time, which takes advantage of virtual machine pausing mechanism to manipulate the virtual machine clock and hide the real I/O completion time. Our approach can be implemented just requiring the hypervisor to be modified, providing a high degree of compatibility and flexibility since it is not necessary to modify neither the operating system nor the application. We evaluate our tool under a real system using old magnetic disks to emulate faster storage devices. Experiments using our technique presented an average latency error of 6.08% for read operations and 6.78% for write operations when comparing a real to device., This work was partially supported by the Spanish Ministry of Science and Innovation under the TIN2015–65316 grant, the Generalitat de Catalunya under contract 2014–SGR–1051., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Approaches for Task Affinity in OpenMP

  • Terboven, Christian
  • Hahnfeld, Jonas
  • Teruel, Xavier
  • Mateo, Sergi
  • Duran, Alejandro
  • Klemm, Michael
  • Olivier, Stephen L.
  • Supinski, Bronis R.
OpenMP tasking supports parallelization of irregular algorithms. Recent OpenMP specifications extended tasking to increase functionality and to support optimizations, for instance with the taskloop construct. However, task scheduling remains opaque, which leads to inconsistent performance on NUMA architectures. We assess design issues for task affinity and explore several approaches to enable it. We evaluate these proposals with implementations in the Nanos++ and LLVM OpenMP runtimes that improve performance up to 40 % and significantly reduce execution time variation., Sandia National Laboratories is a multi-program laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corporation, for the U.S. Department of Energys National Nuclear Security Administration under contract DE-AC04-94AL85000. This work has been developed with the support of the grant SEV-2011-00067 of the Severo Ochoa Program, awarded by the Spanish Government, by the Spanish Ministry of Science and Innovation (TIN2015-65316-P, Computacion de Altas Prestaciones VII) and by the Intel-BSC Exascale Lab collaboration project. Some of the experiments were performed with computing resources granted by JARA- HPC from RWTH Aachen University under project jara0001. Parts of this work were funded by the German Federal Ministry of Research and Education (BMBF) under grant numbers 01IH13008A(ELP). Intel and Xeon are trademarks or registered trademarks of Intel Corporation or its subsidiaries in the United States and other countries. * Other names and brands are the property of their respective owners. Software and workloads used in performance tests may have been optimized for per- formance only on Intel microprocessors. Performance tests, such as SYSmark and Mobile-Mark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more information go to http://www.intel.com/performance. Intel’s compilers may or may not optimize to the same degree for non-Intel micro- processors for optimizations that are not unique to Intel microprocessors. These opti- mizations include SSE2, SSE3, and SSSE3 instruction sets and other optimizations. Intel does not guarantee the availability, functionality, or effectiveness of any optimization on microprocessors not manufactured by Intel. Microprocessor-dependent optimizations in this product are intended for use with Intel microprocessors. Certain optimizations not specific to Intel microarchitecture are reserved for Intel microprocessors. Please refer to the applicable product User and Reference Guides for more information regarding the specific instruction sets covered by this notice., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Optimisation opportunities and evaluation for GPGPU applications on low-end mobile GPUs

  • Trompouki, Matina M.
  • Kosmidis, Leonidas
Previous works in the literature have shown the feasibility of general purpose computations for non-visual applications on low-end mobile graphics processors using graphics APIs. These works focused only on the functional aspects of the software, ignoring the implementation details and therefore their performance implications due to their particular micro-architecture. Since various steps in such applications can be implemented in multiple ways, we identify optimisation opportunities, explore the different options and evaluate them. We show that the implementation details can significantly affect the obtained performance with discrepancies up to 3 orders of magnitude and we demonstrate the effectiveness of our proposal on two embedded platforms, obtaining more than 16× speedup over benchmarks designed following OpenGL ES 2 best practices., This work has been partially supported by the Spanish Ministry of Science and Innovation under grant TIN2015-65316-P and the HiPEAC Network of Excellence., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Towards limiting the impact of timing anomalies in complex real-time processors

  • Benedicte Illescas, Pedro|||0000-0003-1670-7783
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Mezzetti, Enrico|||0000-0002-1886-2931
  • Cazorla Almeida, Francisco Javier
  • Hernández Luz, Carles
Timing verification of embedded critical real-time systems is hindered by complex designs. Timing anomalies, deeply analyzed in static timing analysis, require specific solutions to bound their impact. For the first time, we study the concept and impact of timing anomalies in measurement-based timing analysis, the most used in industry, showing that they require to be considered and handled differently. In addition, we analyze anomalies in the context of Measurement-Based Probabilistic Timing Analysis, which simplifies quantifying their impact., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

HetFS: A heterogeneous file system for everyone

  • Koloventzos, Georgios
  • Nou Castell, Ramon|||0000-0003-0464-1473
  • Miranda, Alberto|||0000-0002-1386-628X
  • Cortés, Toni|||0000-0002-2537-8937
Storage devices have been getting more and more diverse during the last decade. The advent of SSDs made it painfully clear that rotating devices, such as HDDs or magnetic tapes, were lacking in regards to response time. However, SSDs currently have a limited number of write cycles and a significantly larger price per capacity, which has prevented rotational technologies from begin abandoned. Additionally, Non-Volatile Memories (NVMs) have been lately gaining traction, offering devices that typically outperform NAND-based SSDs but exhibit a full new set of idiosyncrasies. Therefore, in order to appropriately support this diversity, intelligent mechanisms will be needed in the near-future to balance the benefits and drawbacks of each storage technology available to a system. In this paper, we present a first step towards such a mechanism called HetFS, an extension to the ZFS file system that is capable of choosing the storage device a file should be kept in according to preprogrammed filters. We introduce the prototype and show some preliminary results of the effects obtained when placing specific files into different devices., The research leading to these results has received funding from the European Community under the BIGStorage ETN (Project 642963 of the H2020-MSCA-ITN-2014), by the Spanish Ministry of Economy and Competitiveness under the TIN2015-65316 grant and by the Catalan Government under the 2014-SGR- 1051 grant. To learn more about the BigStorage project, please visit http: //bigstorage-project.eu/., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

A runtime heuristic to selectively replicate tasks for application-specific reliability targets

  • Subasi, Omer
  • Yalcin, Gulay
  • Zyulkyarov, Ferad
  • Unsal, Osman Sabri
  • Labarta Mancho, Jesús José|||0000-0002-7489-4727
In this paper we propose a runtime-based selective task replication technique for task-parallel high performance computing applications. Our selective task replication technique is automatic and does not require modification/recompilation of OS, compiler or application code. Our heuristic, we call App_FIT, selects tasks to replicate such that the specified reliability target for an application is achieved. In our experimental evaluation, we show that App FIT selective replication heuristic is low-overhead and highly scalable. In addition, results indicate that complete task replication is overkill for achieving reliability targets. We show that with App FIT, we can tolerate pessimistic exascale error rates with only 53% of the tasks being replicated., This work was supported by FI-DGR 2013 scholarship and the European Community’s Seventh Framework Programme [FP7/2007-2013] under the Mont-blanc 2 Project (www.montblanc-project.eu), grant agreement no. 610402 and in part by the European Union (FEDER funds) under contract TIN2015-65316-P., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Improving the integration of task nesting and dependencies in OpenMP

  • Pérez, Josep M.
  • Beltran Querol, Vicenç
  • Labarta Mancho, Jesús José|||0000-0002-7489-4727
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
© 2017 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes,creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works, The tasking model of OpenMP 4.0 supports both nesting and the definition of dependences between sibling tasks. A natural way to parallelize many codes with tasks is to first taskify the high-level functions and then to further refine these tasks with additional subtasks. However, this top-down approach has some drawbacks since combining nesting with dependencies usually requires additional measures to enforce the correct coordination of dependencies across nesting levels. For instance, most non-leaf tasks need to include a taskwait at the end of their code. While these measures enforce the correct order of execution, as a side effect, they also limit the discovery of parallelism. In this paper we extend the OpenMP tasking model to improve the integration of nesting and dependencies. Our proposal builds on both formulas, nesting and dependencies, and benefits from their individual strengths. On one hand, it encourages a top-down approach to parallelizing codes that also enables the parallel instantiation of tasks. On the other hand, it allows the runtime to control dependencies at a fine grain that until now was only possible using a single domain of dependencies. Our proposal is realized through additions to the OpenMP task directive that ensure backward compatibility with current codes. We have implemented a new runtime with these extensions and used it to evaluate the impact on several benchmarks. Our initial findings show that our extensions improve performance in three areas. First, they expose more parallelism. Second, they uncover dependencies across nesting levels, which allows the runtime to make better scheduling decisions. And third, they allow the parallel instantiation of tasks with dependencies between them., This work is supported by the Spanish Government through Programa Severo Ochoa (SEV-2015-0493), by the Spanish Ministry of Science and Technology (project TIN2015-65316-P) and by the Generalitat de Catalunya (grant 2014-SGR-1051)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Assembling a high-productivity DSL for computational fluid dynamics

  • Macià Sorrosal, Sandra
  • Martínez Ferrer, Pedro José|||0000-0002-6097-6676
  • Mateo Bellido, Sergi
  • Beltran Querol, Vicenç
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
“The final publication is available at ACM via http://dx.doi.org/10.1145/3324989.3325721”, As we move towards exascale computing, an abstraction for effective parallel computation is increasingly needed to overcome the maintainability and portability of scientific applications while ensuring the efficient and full exploitation of high-performance systems. These circumstances require computer and domain scientists to work jointly toward a productive working environment. Domain specific languages address this challenge by abstracting the high-level application layer from the final, complex parallel low-level code. Saiph is an innovative domain specific language designed to reduce the work of computational fluid dynamics domain experts to an unambiguous and straightforward transcription of their problem equations. The high-level language, domain-specific compiler and underlying library are enhanced to make applications developed by scientists intuitive. Additions and improvements are presented, designed for the significant advantage of running computational fluid dynamics applications on different machines with no porting or maintenance issues. Numerical methods and parallel strategies are independently added at the library level covering the explicit finite differences resolution of a vast range of problems. Depending on the application, a specific parallel resolution is automatically derived and applied within Saiph, freeing the user from decisions related to numerical methods or parallel executions while ensuring suitable computations. Through a list of benchmarks, we demonstrate the utility and productivity of the Saiph high-level language together with the correctness and performance of the underlying parallel numerical algorithms., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

POSTER: SPiDRE: accelerating sparse memory access patterns

  • Barredo Ferreira, Adrián
  • Beard, Jonathan C.
  • Moretó Planas, Miquel|||0000-0002-9848-8758
Development in process technology has led to an exponential increase in processor speed and memory capacity. However, memory latencies have not improved as dramatically and represent a well-known problem in computer architecture. Cache memories provide more bandwidth with lower latencies than main memories but they are capacity limited. Locality-friendly applications benefit from a large and deep cache hierarchy. Nevertheless, this is a limited solution for applications suffering from sparse and irregular memory access patterns, such as data analytics. In order to accelerate them, we should maximize usable bandwidth, reduce latency and maximize moved data reuse. In this work we explore the Sparse Data Rearrange Engine (SPiDRE), a novel hardware approach to accelerate these applications through near-memory data reorganization., This work has been supported by the Spanish Ministry of Science and Innovation (contract TIN2015-65316-P, Ramon y Cajal fellowship number RYC-2016-21104 and FPI fellowship number BES-2017-080635), and by the Arm-BSC Centre of Excellence initiative., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Modeling high-performance wormhole NoCs for critical real-time embedded systems

  • Panic, Milos
  • Hernández, Carles
  • Quiñones, Eduardo
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Cazorla Almeida, Francisco Javier
Manycore chips are a promising computing platform to cope with the increasing performance needs of critical real-time embedded systems (CRTES). However, manycores adoption by CRTES industry requires understanding task's timing behavior when their requests use manycore's network-on-chip (NoC) to access hardware shared resources. This paper analyzes the contention in wormhole-based NoC (wNoC) designs - widely implemented in the high-performance domain - for which we introduce a new metric: worst-contention delay (WCD) that captures wNoC impact on worst-case execution time (WCET) in a tighter manner than the existing metric, worst-case traversal time (WCTT). Moreover, we provide an analytical model of the WCD that requests can suffer in a wNoC and we validate it against wNoC designs resembling those in the Tilera-Gx36 and the Intel-SCC 48-core processors. Building on top of our WCD analytical model, we analyze the impact on WCD that different design parameters such as the number of virtual channels, and we make a set of recommendations on what wNoC setups to use in the context of CRTES., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

An architecture model for a distributed virtualization system

  • Pessolani, Pablo
  • Tinetti, Fernando
  • Cortés, Toni|||0000-0002-2537-8937
  • Gonnet, Silvio
This article presents an architecture model for a Distributed Virtualization System, which could expand a virtual execution environment from a single physical machine to several nodes of a cluster. With current virtualization technologies, computing power and resource usage of Virtual Machines (or Containers) are limited to the physical machine where they run. To deliver high levels of performance and scalability, cloud applications are usually partitioned in several Virtual Machines (or Containers) located on different nodes of a virtualization cluster. Developers often use that processing model because the same instance of the operating system is not available on each node where their components run. The proposed architecture model is suitable for new trends in software development because it is inherently distributed. It combines and integrates Virtualization and Distributed Operating Systems technologies with the benefits of both worlds, providing the same isolated instance of a Virtual Operating System on each cluster node. Although it requires the introduction of changes in existing operating systems, thousands of legacy applications would not require modifications to obtain their benefits. A Distributed Virtualization System is suitable to deliver high-performance cloud services with provider-class features, such as high-availability, replication, migration, and load balancing. Furthermore, it is able to concurrently run several isolated instances of different guest Virtual Operating Systems, allocating a subset of nodes for each instance and sharing nodes between them. Currently, a prototype is running on a cluster of commodity hardware provided with two kinds of Virtual Operating Systems tailored for internet services (web server) as a proof of concept., Toni Cortes’ involvement in this work was financially supported by the Spanish National Government (financial grant SEV2015-0493 of Severo Ochoa program), the Spanish Ministry of Science and Innovation (contract TIN2015-65316), and the Government of Catalonia (contract 2014-SGR-1051).Fernando G. Tinetti’s involvement in this work was financially supported by Universidad de La Plata (Faculty of Informatics) and the Scientific Research Board (CIC, for its Spanish initials) of Buenos Aires, Argentina., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Micro-architectural characterization of Apache Spark on batch and stream processing workloads

  • Awan, Ahsan
  • Brorsson, Mats
  • Vlassov, Vladimir
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
While cluster computing frameworks are continuously evolving to provide real-time data analysis capabilities, Apache Spark has managed to be at the forefront of big data analytics for being a unified framework for both, batch and stream data processing. However, recent studies on micro-architectural characterization of in-memory data analytics are limited to only batch processing workloads. We compare the micro-architectural performance of batch processing and stream processing workloads in Apache Spark using hardware performance counters on a dual socket server. In our evaluation experiments, we have found that batch processing and stream processing has same micro-architectural behavior in Spark if the difference between two implementations is of micro-batching only. If the input data rates are small, stream processing workloads are front-end bound. However, the front end bound stalls are reduced at larger input data rates and instruction retirement is improved. Moreover, Spark workloads using DataFrames have improved instruction retirement over workloads using RDDs., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Cache side-channel attacks and time-predictability in high-performance critical real-time systems

  • Trilla Rodríguez, David
  • Hernández Luz, Carles
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Cazorla Almeida, Francisco Javier
Embedded computers control an increasing number of systems directly interacting with humans, while also manage more and more personal or sensitive information. As a result, both safety and security are becoming ubiquitous requirements in embedded computers, and automotive is not an exception to that. In this paper we analyze time-predictability (as an example of safety concern) and side-channel attacks (as an example of security issue) in cache memories. While injecting randomization in cache timing-behavior addresses each of those concerns separately, we show that randomization solutions for time-predictability do not protect against side-channel attacks and vice-versa. We then propose a randomization solution to achieve both safety and security goals., This work has been partially funded by the Spanish Ministry of Science and Innovation under grant TIN2015-65316-P. Jaume Abella has been partially supported by the Ministry of Economy and Competitiveness under Ramon y Cajal fellowship number RYC-2013-14717. Authors want to thank Boris Kpf for his technical comments in early versions of this work., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Improving accuracy and speeding up document image classification through parallel systems

  • Ferrando Monsonís, Javier|||0000-0002-2637-0961
  • Domínguez, Juan Luis
  • Torres Viñals, Jordi|||0000-0003-1963-7418
  • García Fuentes, Raul
  • García Doménech, David
  • Garrido Miñambres, Daniel
  • Cortada, Jordi
  • Valero Cortés, Mateo|||0000-0003-2917-2482
This paper presents a study showing the benefits of the EfficientNet models compared with heavier Convolutional Neural Networks (CNNs) in the Document Classification task, essential problem in the digitalization process of institutions. We show in the RVL-CDIP dataset that we can improve previous results with a much lighter model and present its transfer learning capabilities on a smaller in-domain dataset such as Tobacco3482. Moreover, we present an ensemble pipeline which is able to boost solely image input by combining image model predictions with the ones generated by BERT model on extracted text by OCR. We also show that the batch size can be effectively increased without hindering its accuracy so that the training process can be sped up by parallelizing throughout multiple GPUs, decreasing the computational time needed. Lastly, we expose the training performance differences between PyTorch and Tensorflow Deep Learning frameworks., This work was partially supported by the Spanish Ministry of Science and Innovation and the European Regional Development Fund under contract TIN2015- 65316-P, by the BSC-CNS Severo Ochoa program SEV-2015-0493, and grant 2017-SGR-1414 by Generalitat de Catalunya and by the research agreement CaixaBank-BSC 2016-2021., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Improving OpenStack Swift interaction with the I/O stack to enable software defined storage

  • Nou Castell, Ramon|||0000-0003-0464-1473
  • Miranda, Alberto|||0000-0002-1386-628X
  • Siquier, Marc
  • Cortés, Toni|||0000-0002-2537-8937
This paper analyses how OpenStack Swift, a distributed object storage service for a globally used middleware, interacts with the I/O subsystem through the Operating System. This interaction, which seems organised and clean on the middleware side, becomes disordered on the device side when using mechanical disk drives, due to the way threads are used internally to request data. We will show that only modifying the Swift threading model we achieve an 18% mean improvement in performance with objects larger than 512 KiB and obtain a similar performance with smaller objects. Compared to the original scenario, the performance obtained on both scenarios is obtained in a fair way: the bandwidth is shared equally between concurrently accessed objects. Moreover, this threading model allows us to apply techniques for Software Defined Storage (SDS). We show an implementation of a Bandwidth Differentiation technique that can control each data stream and that guarantees a high utilization of the device., The research leading to these results has received funding from the European Community under the IOStack (H2020-ICT-2014-7-1) project, by the Spanish Ministry of Economy and Competitiveness under the TIN2015-65316-P grant and by the Catalan Government under the 2014-SGR-1051 grant. To learn more about the IOStack H2020 project, please visit http:nnwww.iostack.eu., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

MUSA: a multi-level simulation approach for next-generation HPC machines

  • Grass, Thomas
  • Allande, César
  • Armejach, Adrià
  • Rico, Alejandro
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
  • Labarta Mancho, Jesús José|||0000-0002-7489-4727
  • Valero Cortés, Mateo|||0000-0003-2917-2482
  • Casas, Marc|||0000-0003-4564-2093
  • Moretó Planas, Miquel|||0000-0002-9848-8758
The complexity of High Performance Computing (HPC) systems is increasing in the number of components and their heterogeneity. Interactions between software and hardware involve many different aspects which are typically not transparent to scientific programmers and system architects. Therefore, predicting the behavior of current scientific applications on future HPC infrastructures is a challenging task. In this paper we present MUSA, an end-to-end methodology that employs a multi-level simulation infrastructure. By combining different levels of abstraction, MUSA is able to model the communication network, microarchitectural details and system software interactions, providing different trade-offs in terms of simulation cost and accuracy. We compare detailed MUSA simulations with native executions of up to 2,048 cores and find relative errors that are within 10% in the common case. In addition, we use MUSA to simulate up to 16,384 cores and successfully identify scalability bottlenecks due to different factors, e.g. memory contention or load imbalance. We also compare different system configurations, showing how MUSA can help system designers to assess the usefulness of future technologies in next-generation HPC machines., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Interconnect Energy Savings and Lower Latency Networks in Hadoop Clusters: The Missing Link

  • Fischer e Silva, Renan
  • Carpenter, Paul Matthew
An important challenge of modern data centres running Hadoop workloads is to minimise energy consumption, a significant proportion of which is due to the network. Significant network savings are already possible using Energy Efficient Ethernet, supported by a large number of NICs and switches, but recent work has demonstrated that the packet coalescing settings must be carefully configured to avoid a substantial loss in performance. Meanwhile, Hadoop is evolving from its original batch concept to become a more iterative type of framework. Other recent work attempts to reduce Hadoop's network latency using Explicit Congestion Notifications. Linking these studies reveals that, surprisingly, even when packet coalescing does not hurt performance, it can degrade network latency much more than previously thought. This paper is the first to analyze the impact of packet coalescing in the context of network latency. We investigate how to design and configure interconnects to provide the maximum energy savings without degrading cluster throughput performance or network latency., The research leading to these results has received funding from the European Unions Seventh Framework Programme (FP7/2007–2013) under grant agreement number 610456 (Euroserver). The research was also supported by the Ministry of Economy and Competitiveness of Spain under the contracts TIN2012-34557 and TIN2015-65316-P, Generalitat de Catalunya (contracts 2014-SGR-1051 and 2014-SGR-1272), HiPEAC-3 Network of Excellence (ICT- 287759), and the Severo Ochoa Program (SEV-2011-00067) of the Spanish Government., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

BRASIL: A high-integrity GPGPU toolchain for automotive systems

  • Trompouki, Matina Maria
  • Kosmidis, Leonidas
Embedded General Purpose Graphics Processing Units (GPGPUs) are increasingly used in automotive to enable Advanced Driving Assistance (ADAS) and Autonomous driving. However, their functional safety certification has been addressed so far only at the application or hardware level. In this work, we demonstrate how the entire GPGPU software stack can be developed and certified up to ASIL-D, including its toolchain. We introduce BRASIL, a GPGPU toolchain targeting high-integrity GPGPU computing and we explain our methodology to achieve ISO 26262 compliance, as well as our tool qualification strategy, This work has been partially supported by the Spanish Ministry of Science and Innovation under grant TIN2015-65316-P and the HiPEAC Network of Excellence. Leonidas Kosmidis is also funded by the Spanish Ministry of Economy and Competitiveness (MINECO) under a Juan de la Cierva Formacion postdoctoral fellowship (FJCI-2017-34095)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Evaluating the effect of last-level cache sharing on integrated GPU-CPU systems with heterogeneous applications

  • García Flores, Víctor
  • Gomez Luna, J.
  • Grass, Thomas Dieter
  • Rico, Alejandro
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
  • Pena, A. J.
Heterogeneous systems are ubiquitous in the field of High- Performance Computing (HPC). Graphics processing units (GPUs) are widely used as accelerators for their enormous computing potential and energy efficiency; furthermore, on-die integration of GPUs and general-purpose cores (CPUs) enables unified virtual address spaces and seamless sharing of data structures, improving programmability and softening the entry barrier for heterogeneous programming. Although on-die GPU integration seems to be the trend among the major microprocessor manufacturers, there are still many open questions regarding the architectural design of these systems. This paper is a step forward towards understanding the effect of on-chip resource sharing between GPU and CPU cores, and in particular, of the impact of last-level cache (LLC) sharing in heterogeneous computations. To this end, we analyze the behavior of a variety of heterogeneous GPU-CPU benchmarks on different cache configurations. We perform an evaluation of the popular Rodinia benchmark suite modified to leverage the unified memory address space. We find such GPGPU workloads to be mostly insensitive to changes in the cache hierarchy due to the limited interaction and data sharing between GPU and CPU. We then evaluate a set of heterogeneous benchmarks specifically designed to take advantage of the finegrained data sharing and low-overhead synchronization between GPU and CPU cores that these integrated architectures enable. We show how these algorithms are more sensitive to the design of the cache hierarchy, and find that when GPU and CPU share the LLC execution times are reduced by 25% on average, and energy-to-solution by over 20% for all benchmarks., This work has been supported by the Spanish Ministry of Science and Innovation (contract TIN2015-65316-P) and by the BSC/UPC NVIDIA GPU Center of Excellence., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

STT-MRAM for real-time embedded systems: performance and WCET implications

  • Asifuzzaman, Kazi
  • Fernández, Mikel
  • Radojković, Petar
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Cazorla Almeida, Francisco Javier
STT-MRAM is an emerging non-volatile memory quickly approaching DRAM in terms of capacity, frequency and device size. Intensified efforts in STT-MRAM research by the memory manufacturers may indicate a revolution with STT-MRAM memory technology is imminent, and therefore it is essential to perform system level research to explore use-cases and identify computing domains that could benefit from this technology. Special STT-MRAM features such as intrinsic radiation hardness, non-volatility, zero stand-by power and capability to function in extreme temperatures makes it particularly suitable for aerospace, avionics and automotive applications. Such applications often have real-time requirements --- that is, certain tasks must complete within a strict deadline. Analyzing whether this deadline is met requires Worst Case Execution Time (WCET) Analysis, which is a fundamental part of evaluating any real-time system. In this study, we investigate the feasibility of using STT-MRAM in real-time embedded systems by analyzing average system performance impact and WCET implications., This work was supported by BSC, Spanish Government through Programa Severo Ochoa (SEV-2015-0493), by the Spanish Ministry of Science and Technology through TIN2015-65316-P project and by the Generalitat de Catalunya (contracts 2014-SGR-1051 and 2014-SGR-1272). This work has also received funding from the European Union’s Horizon 2020 research and innovation programme under ExaNoDe project (grant agreement No 671578). Jaume Abella was partially supported by the Ministry of Economy and Competitive-ness under Ramon y Cajal postdoctoral fellowship RYC-2013-14717., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

VIA: A smart scratchpad for vector units with application to sparse matrix computations

  • Pavón Rivera, Julián
  • Vargas Valdivieso, Iván
  • Barredo Ferreira, Adrián
  • Marimon Illana, Joan
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Moll Echeto, Francisco de Borja|||0000-0002-1290-3253
  • Unsal, Osman Sabri
  • Valero Cortés, Mateo|||0000-0003-2917-2482
  • Cristal Kestelman, Adrián|||0000-0003-1277-9296
Sparse matrix operations are critical kernels in multiple application domains such as High Performance Computing, artificial intelligence and big data. Vector processing is widely used to improve performance on mathematical kernels with dense matrices. Unfortunately, existing vector architectures do not cope well with sparse matrix computations, achieving much lower performance in comparison with their dense counterparts.To overcome this limitation, we present the Vector Indexed Architecture (VIA), a novel hardware vector architecture that accelerates applications with irregular memory access patterns such as sparse matrix computations. There are two main bottlenecks when computing with sparse matrices: irregular memory accesses and index matching. VIA addresses these two bottlenecks with a smart scratchpad that is tightly coupled to the Vector Functional Units within the core.Thanks to this structure, VIA improves locality for sparse-dense computations and improves the index matching search process for sparse computations. As a result, VIA achieves significant performance speedup over highly optimized state-of-the-art C++ algebra libraries. On average, VIA outperforms sparse matrix vector multiplication, sparse matrix addition and sparse matrix matrix multiplication kernels by 4.22 ×, 6.14 × and 6.00 ×, respectively, when evaluated over a thousand sparse matrices that arise in real applications. In addition, we prove the generality of VIA by showing that it can accelerate histogram and stencil applications by 4.5 × and 3.5 ×, respectively., This work has been supported by the European HiPEAC Network of Excellence, by the Spanish Ministry of Science and Innovation (contract TIN2015-65316-P), by the Generalitat de Catalunya (contracts 2017-SGR-1414 and 2017-SGR-1328), and by the DRAC project, which is co-financed by the European Union Regional Development Fund within the framework of the ERDF Operational Program of Catalonia 2014-2020 with a grant of 50% of total eligible cost. A. Barredo has been supported by the Spanish Government under Formation del Personal Investigador fellowship number BES-2017- 080635. M. Moreto has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramon y Cajal fellowship number RYC-2016-21104., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Graph partitioning applied to DAG scheduling to reduce NUMA effects

  • Sánchez Barrera, Isaac|||0000-0003-1616-0685
  • Casas, Marc|||0000-0003-4564-2093
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
  • Labarta Mancho, Jesús José|||0000-0002-7489-4727
  • Valero Cortés, Mateo|||0000-0003-2917-2482
The complexity of shared memory systems is becoming more relevant as the number of memory domains increases, with different access latencies and bandwidth rates depending on the proximity between the cores and the devices containing the data. In this context, techniques to manage and mitigate non-uniform memory access (NUMA) effects consist in migrating threads, memory pages or both and are typically applied by the system software. We propose techniques at the runtime system level to reduce NUMA effects on parallel applications. We leverage runtime system metadata in terms of a task dependency graph. Our approach, based on graph partitioning methods, is able to provide parallel performance improvements of 1.12X on average with respect to the state-of-the-art., This work has been partially supported by the RoMoL ERC Advanced Grant (GA 321253), the European HiPEAC Network of Excellence and the Spanish Government (contract TIN2015-65316-P). I. Sánchez Barrera has been supported by the Spanish Government under Formación del Profesorado Universitario fellowship number FPU15/03612., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Computational Fluid and Particle Dynamics Simulations for Respiratory System: Runtime Optimization on an Arm Cluster

  • Garcia-Gasulla, Marta
  • Josep-Fabrego, Marc
  • Eguzkitza, Beatriz|||0000-0002-3302-6667
  • Mantovani, Filippo|||0000-0003-3559-4825
Computational fluid and particle dynamics simulations (CFPD) are of paramount importance for studying and improving drug effectiveness. Computational requirements of CFPD codes involves high-performance computing (HPC) resources. For these reasons we introduce and evaluate in this paper system software techniques for improving performance and tolerate load imbalance on a state-of-the-art production CFPD code. We demonstrate benefits of these techniques on both Intel- and Arm-based HPC clusters showing the importance of using mechanisms applied at runtime to improve the performance independently of the underlying architecture. We run a real CFPD simulation of particle tracking on the human respiratory system, showing performance improvements of up to 2X, keeping the computational resources constant., This work is partially supported by the Spanish Government (SEV-2015-0493), by the Spanish Ministry of Science and Technology project (TIN2015-65316-P), by the Generalitat de Catalunya (2017-SGR-1414), and by the European Mont-Blanc projects (288777, 610402 and 671697)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

On assessing the viability of probabilistic scheduling with dependent tasks

  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Mezzetti, Enrico|||0000-0002-1886-2931
  • Cazorla, Francisco J.
Despite the significant interest, in the last years, in probabilistic scheduling and probabilistic timing analysis, the interrelation between them has been scarcely addressed. Probabilistic scheduling approaches typically build on a series of assumptions on the probabilistic behavior of each task - or single jobs activations - that have not been shown to be entirely fulfilled by the distributions computed with probabilistic timing analysis. This paper aims at providing a clear understanding of probabilistic Worst-Case Execution Time distributions (pWCET) as a common concept of probabilistic timing and schedulability analysis. We focus on independence of pWCET estimates as the main concern in the application of probabilistic scheduling, with particular emphasis on measurement-based probabilistic timing analyses, for which independence across pWCET estimates may not be guaranteed. We relate pWCET (in)dependence to the platform-induced timing dependencies that occur among tasks, and even jobs of the same task. We conclude that independent pWCET distributions can be obtained, even if dependencies exist, by either controlling the measurement protocol, or by deriving distinct pWCET estimates for particular instances of a task., This work has been partially supported by the Spanish Ministry of Economy and Competitiveness (MINECO) under grant TIN2015-65316-P, the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No. 772773) and the HiPEAC Network of Excellence. Jaume Abella and Enrico Mezzetti have been partially supported by MINECO under Ramon y Cajal and Juan de la Cierva-Incorporación postdoctoral fellowships number RYC-2013-14717 and IJCI-2016-27396 respectively., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Beyond the socket: NUMA-aware GPUs

  • Ugljesa, Milic
  • Villa, Oreste
  • Bolotin, Evgeny
  • Arunkumar, Akhil
  • Ebrahimi, Eiman
  • Jaleel, Aamer
  • Ramirez, Alex
  • Nellans, David
GPUs achieve high throughput and power efficiency by employing many small single instruction multiple thread (SIMT) cores. To minimize scheduling logic and performance variance they utilize a uniform memory system and leverage strong data parallelism exposed via the programming model. With Moore's law slowing, for GPUs to continue scaling performance (which largely depends on SIMT core count) they are likely to embrace multi-socket designs where transistors are more readily available. However when moving to such designs, maintaining the illusion of a uniform memory system is increasingly difficult. In this work we investigate multi-socket non-uniform memory access (NUMA) GPU designs and show that significant changes are needed to both the GPU interconnect and cache architectures to achieve performance scalability. We show that application phase effects can be exploited allowing GPU sockets to dynamically optimize their individual interconnect and cache policies, minimizing the impact of NUMA effects. Our NUMA-aware GPU outperforms a single GPU by 1.5×, 2.3×, and 3.2× while achieving 89%, 84%, and 76% of theoretical application scalability in 2, 4, and 8 sockets designs respectively. Implementable today, NUMA-aware multi-socket GPUs may be a promising candidate for scaling GPU performance beyond a single socket., We would like to thank anonymous reviewers and Steve Keckler for their help in improving this paper. The first author is supported by the Ministry of Economy and Competitiveness of Spain (TIN2012-34557, TIN2015-65316-P, and BES-2013-063925), Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Rebalancing the core front-end through HPC code analysis

  • Milic, Ugljesa
  • Carpenter, Paul Matthew
  • Rico, Alejandro
  • Ramirez, Alex
There is a need to increase performance under the same power and area envelope to achieve Exascale technology in high performance computing (HPC). The today's chip multiprocessor (CMP) design is tailored by traditional desktop and server workloads, different from parallel applications commonly run in HPC. In this work, we focus on the HPC code characteristics and processor front-end which factors around 30% of core power and area on the emerging lean-core type of processors used in HPC. Separating serial from parallel code sections inside applications, we characterize three HPC benchmark suites and compare them to a traditional set of desktop integer workloads. HPC applications have biased and mostly backward taken branches, small dynamic instruction footprints, and long basic blocks. Our findings suggest smaller branch predictors (BP) with the additional loop BP, smaller branch target buffers (BTB), and smaller L1 instruction caches (I-cache) with wider lines. Still, the aforementioned downsizing applies only to the cores meant to run parallel code. The difference between serial and parallel code sections in HPC applications points to an asymmetric CMP design, with one baseline core for sequential and many HPCtailored cores designed for parallel code. Predictions using Sniper simulator and McPAT show that an HPC-tailored lean core saves 16% of the core area and 7% of power compared to a baseline core, without performance loss. Using the area savings to add an extra core, an asymmetric CMP with one baseline and eight tailored cores has the same area budget as a symmetric CMP composed out of eight baseline cores demanding 4% more power and providing 12% shorter execution time on average.
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Full-network embedding in a multimodal embedding pipeline

  • Vilalta Arias, Armand
  • Garcia Gasulla, Dario|||0000-0001-6732-5641
  • Parés Pont, Ferran
  • Moreno Vázquez, Jonatan
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
  • Labarta Mancho, Jesús José|||0000-0002-7489-4727
  • Cortés García, Claudio Ulises|||0000-0003-0192-3096
  • Suzumura, Toyotaro
The current state-of-the-art for image annotation and image retrieval tasks is obtained through deep neural networks, which combine an image representation and a text representation into a shared embedding space. In this paper we evaluate the impact of using the Full-Network embedding in this setting, replacing the original image representation in a competitive multimodal embedding generation scheme. Unlike the one-layer image embeddings typically used by most approaches, the Full-Network embedding provides a multi-scale representation of images, which results in richer characterizations. To measure the influence of the Full-Network embedding, we evaluate its performance on three different datasets, and compare the results with the original multimodal embedding generation scheme when using a one-layer image embedding, and with the rest of the state-of-the-art. Results for image annotation and image retrieval tasks indicate that the Full-Network embedding is consistently superior to the one-layer embedding. These results motivate the integration of the Full-Network embedding on any multimodal embedding generation scheme, something feasible thanks to the flexibility of the approach, This work is partially supported by the Joint Study Agreement no. W156463 under the IBM/BSC Deep Learning Center agreement, by the Spanish Government through Programa Severo Ochoa (SEV-2015- 0493), by the Spanish Ministry of Science and Technology through TIN2015-65316-P project, by the Generalitat de Catalunya (contracts 2014-SGR-1051), and by the Core Research for Evolutional Science and Technology (CREST) program of Japan Science and Technology Agency (JST)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Holistic slowdown driven scheduling and resource management for malleable jobs

  • D'Amico, Marco|||0000-0002-6195-6204
  • Jokanovic, Ana
  • Corbalán González, Julita|||0000-0002-3926-5634
In job scheduling, the concept of malleability has been explored since many years ago. Research shows that malleability improves system performance, but its utilization in HPC never became widespread. The causes are the difficulty in developing malleable applications, and the lack of support and integration of the different layers of the HPC software stack. However, in the last years, malleability in job scheduling is becoming more critical because of the increasing complexity of hardware and workloads. In this context, using nodes in an exclusive mode is not always the most efficient solution as in traditional HPC jobs, where applications were highly tuned for static allocations, but offering zero flexibility to dynamic executions. This paper proposes a new holistic, dynamic job scheduling policy, Slowdown Driven (SD-Policy), which exploits the malleability of applications as the key technology to reduce the average slowdown and response time of jobs. SD-Policy is based on backfill and node sharing. It applies malleability to running jobs to make room for jobs that will run with a reduced set of resources, only when the estimated slowdown improves over the static approach. We implemented SD-Policy in SLURM and evaluated it in a real production environment, and with a simulator using workloads of up to 198K jobs. Results show better resource utilization with the reduction of makespan, response time, slowdown, and energy consumption, up to respectively 7%, 50%, 70%, and 6%, for the evaluated workloads., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Future vector microprocessor extensions for data aggregations

  • Hayes, Timothy
  • Palomar, Oscar
  • Unsal, Osman Sabri
  • Cristal Kestelman, Adrián|||0000-0003-1277-9296
  • Valero Cortés, Mateo|||0000-0003-2917-2482
As the rate of annual data generation grows exponentially, there is a demand to aggregate and summarise vast amounts of information quickly. In the past, frequency scaling was relied upon to push application throughput. Today, Dennard scaling has ceased and further performance must come from exploiting parallelism. Single instruction-multiple data (SIMD) instruction sets offer a highly efficient and scalable way of exploiting data-level parallelism (DLP). While microprocessors originally offered very simple SIMD support targeted at multimedia applications, these extensions have been growing both in width and functionality. Observing this trend, we use a simulation framework to model future SIMD support and then propose and evaluate five different ways of vectorising data aggregation. We find that although data aggregation is abundant in DLP, it is often too irregular to be expressed efficiently using typical SIMD instructions. Based on this observation, we propose a set of novel algorithms and SIMD instructions to better capture this irregular DLP. Furthermore, we discover that the best algorithm is highly dependent on the characteristics of the input. Our proposed solution can dynamically choose the optimal algorithm in the majority of cases and achieves speedups between 2.7x and 7.6x over a scalar baseline., The research leading to these results has received funding from the RoMoL ERC Advanced Grant GA no 321253 and is supported in part by the European Union (FEDER funds) under contract TTIN2015-65316-P. Timothy Hayes is supported by a FPU research grant from the Spanish MECD., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

TaskPoint: sampled simulation of task-based programs

  • Grass, Thomas Dieter
  • Rico, Alejandro
  • Casas, Marc|||0000-0003-4564-2093
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
Sampled simulation is a mature technique for reducing simulation time of single-threaded programs, but it is not directly applicable to simulation of multi-threaded architectures. Recent multi-threaded sampling techniques assume that the workload assigned to each thread does not change across multiple executions of a program. This assumption does not hold for dynamically scheduled task-based programming models. Task-based programming models allow the programmer to specify program segments as tasks which are instantiated many times and scheduled dynamically to available threads. Due to system noise and variation in scheduling decisions, two consecutive executions on the same machine typically result in different instruction streams processed by each thread. In this paper, we propose TaskPoint, a sampled simulation technique for dynamically scheduled task-based programs. We leverage task instances as sampling units and simulate only a fraction of all task instances in detail. Between detailed simulation intervals we employ a novel fast-forward mechanism for dynamically scheduled programs. We evaluate the proposed technique on a set of 19 task-based parallel benchmarks and two different architectures. Compared to detailed simulation, TaskPoint accelerates architectural simulation with 64 simulated threads by an average factor of 19.1 at an average error of 1.8% and a maximum error of 15.0%., This work has been supported by the Spanish Government (Severo Ochoa grants SEV2015-0493, SEV-2011-00067), the Spanish Ministry of Science and Innovation (contract TIN2015-65316-P), Generalitat de Catalunya (contracts 2014-SGR-1051 and 2014-SGR-1272), the RoMoL ERC Advanced Grant (GA 321253), the European HiPEAC Network of Excellence and the Mont-Blanc project (EU-FP7-610402 and EU-H2020-671697). M. Moreto has been partially supported by the Ministry of Economy and Competitiveness under Juan de la Cierva postdoctoral fellowship JCI-2012-15047. M. Casas is supported by the Ministry of Economy and Knowledge of the Government of Catalonia and the Cofund programme of the Marie Curie Actions of the EUFP7 (contract 2013BP B 00243). T.Grass has been partially supported by the AGAUR of the Generalitat de Catalunya (grant 2013FI B 0058)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Improving predication efficiency through compaction/restoration of SIMD instructions

  • Barredo Ferreira, Adrián
  • Cebrián González, Juan Manuel
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Casas, Marc|||0000-0003-4564-2093
  • Valero Cortés, Mateo|||0000-0003-2917-2482
Vector processors offer a wide range of unexplored opportunities to improve performance and energy efficiency. However, despite its potential, vector code generation and execution have significant challenges, the most relevant ones being control flow divergence. Most modern processors including SIMD extensions (such as AVX) rely on predication to support divergence control. In predicated codes, performance and energy consumption are usually insensitive to the number of true values in a predicated mask. This implies that the system efficiency becomes sub-optimal as vector length increases. In this paper we focus on SIMD extensions and propose a novel approach to improve execution efficiency in predicated SIMD instructions, the Compaction/Restoration (CR) technique. CR delays predicated SIMD instructions with inactive elements and compacts them with instances of the same instruction from different loop iterations to form an equivalent dense vector instruction, where, in the best case, all the elements are active. After executing such dense instructions, their results are restored to the original instructions. Our evaluation shows that CR improves performance by up to 25% and reduces dynamic energy consumption by up to 43% on real unmodified applications with predicated execution. Moreover, CR allows executing unmodified legacy code with short vector instructions (AVX-2) on newer architectures with wider vectors (AVX-512), achieving up to 56% performance benefits., This work has been partially supported by the RoMoL ERC Advanced Grant (GA 321253), the European HiPEAC Network of Excellence, the Spanish Government (contract TIN2015-65316-P) and the European Union’s Horizon 2020 research and innovation program under the Mont-Blanc 2020 project (grant agreement 779877). A. Barredo has been supported by the Spanish Government under Formación del Personal Investigador fellowship number BES-2017-080635. M. Moretó and M. Casas have been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramon y Cajal fellowship numbers RYC-2016-21104 and RYC-2017-23269., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Evaluating SLURM simulator with real-machine SLURM and vice versa

  • Jokanovic, Ana
  • D'Amico, Marco|||0000-0002-6195-6204
  • Corbalán González, Julita|||0000-0002-3926-5634
© 2018 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes,creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works., Having a precise and a fast job scheduler model that resembles the real-machine job scheduling software behavior is extremely important in the field of job scheduling. The idea behind SLURM simulator is preserving the original code of the core SLURM functions while allowing for all the advantages of a simulator. Since 2011, SLURM simulator has passed through several iterations of improvements in different research centers. In this work, we present our latest improvements of SLURM simulator and perform the first-ever validation of the simulator on the real machine. In particular, we improved the simulator's performance for about 2.6 times, made the simulator deterministic across several same set-up runs, and improved the simulator's accuracy; its deviation from the real-machine is lowered from previous 12% to at most 1.7%. Finally, we illustrate with several use cases the value of the simulator for job scheduling researchers, SLURM-system administrators, and SLURM developers., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Performance meets programmabilty: Enabling native Python MPI tasks in PyCOMPSs

  • Elshazly, Hatem Mohamed Abdelfattah Eid|||0000-0002-8591-1502
  • Lordan Gomis, Francesc|||0000-0002-9845-8890
  • Ejarque Artigas, Jorge
  • Badia Sala, Rosa Maria|||0000-0003-2941-5499
©2020 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes,creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works., The increasing complexity of modern and future computing systems makes it challenging to develop applications that aim for maximum performance. Hybrid parallel programming models offer new ways to exploit the capabilities of the underlying infrastructure. However, the performance gain is sometimes accompanied by increased programming complexity. We introduce an extension to PyCOMPSs, a high-level task-based parallel programming model for Python applications, to support tasks that use MPI natively as part of the task model. Without compromising application's programmability, using Native MPI tasks in PyCOMPSs offers up to 3x improvement in total performance for compute intensive applications and up to 1.9x improvement in total performance for I/O intensive applications over sequential implementation of the tasks., This work is partially supported by the European Union through the Horizon 2020 research and innovation programme under contracts 721865 (EXPERTISE Project) and 800898 (ExaQUte project), by the Spanish Government (TIN2015- 65316-P) and the Generalitat de Catalunya (contract 2014- SGR-1051)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Accelerating K-mer Frequency Counting with GPU and Non-Volatile Memory

  • Cadenelli, Nicola
  • Polo Bardés, Jordà
  • Carrera, David
The emergence of Next Generation Sequencing (NGS) platforms has increased the throughput of genomic sequencing and in turn the amount of data that needs to be processed, requiring highly efficient computation for its analysis. In this context, modern architectures including accelerators and non-volatile memory are essential to enable the mass exploitation of these bioinformatics workloads. This paper presents a redesign of the main component of a state-of-the-art reference-free method for variant calling, SMUFIN, which has been adapted to make the most of GPUs and NVM devices. SMUFIN relies on counting the frequency of k-mers (substrings of length k) in DNA sequences, which also constitutes a well-known problem for many bioinformatics workloads, such as genome assembly. We propose techniques to improve the efficiency of k-mer counting and to scale-up workloads like SMUFIN that used to require 16 nodes of Marenostrum 3 to a single machine with a GPU and NVM drives. Results show that although the single machine is not able to improve the time to solution of 16 nodes, its CPU time is 7.5x shorter than the aggregate CPU time of the 16 nodes, with a reduction in energy consumption of 5.5x., This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 639595). It is also partially supported by the Ministry of Economy of Spain under contract TIN2015-65316-P and Generalitat de Catalunya under contract 2014SGR1051, by the ICREA Academia program, and by the BSC-CNS Severo Ochoa program (SEV-2015-0493). We are also grateful to SandDisk for lending the FusionIO cards and to Nvidia who donated the Tesla K40c., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

An approach for detecting power peaks during testing and breaking systematic pathological behavior

  • Trilla Rodríguez, David
  • Hernández Luz, Carles
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Cazorla Almeida, Francisco Javier
The verification and validation process of embedded critical systems requires providing evidence of their functional correctness and also that their non-functional behavior stays within limits. In this work, we focus on power peaks, which may cause voltage droops and thus, challenge performance to preserve correct operation upon droops. In this line, the use of complex software and hardware in critical embedded systems jeopardizes the confidence that can be placed on the tests carried out during the campaigns performed at analysis. This is so because it is unknown whether tests have triggered the highest power peaks that can occur during operation and whether any such peak can occur systematically. In this paper we propose the use of randomization, already used for timing analysis of real-time systems, as an enabler to guarantee that (1) tests expose those peaks that can arise during operation and (2) peaks cannot occur systematically inadvertently., This work has been partially supported by the Spanish Ministry of Economy and Competitiveness (MINECO) under grant TIN2015-65316-P, the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No. 772773), and the HiPEAC Network of Excellence. MINECO partially supported Jaume Abella under Ramon y Cajal postdoctoral fellowship (RYC-2013-14717)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Implementation of the K-Means Algorithm on Heterogeneous Devices: A Use Case Based on an Industrial Dataset

  • Xu, Ying hao
  • Vidal, Miquel|||0000-0002-1973-8289
  • Arejita, Beñat
  • Diaz, Javier
  • Alvarez, Carlos
  • Jiménez González, Daniel|||0000-0001-6064-7883
  • Martorell Bofill, Xavier|||0000-0002-0417-3430
  • Mantovani, Filippo|||0000-0003-3559-4825
This paper presents and analyzes a heterogeneous implementation of an industrial use case based on K-means that targets symmetric multiprocessing (SMP), GPUs and FPGAs. We present how the application can be optimized from an algorithmic point of view and how this optimization performs on two heterogeneous platforms. The presented implementation relies on the OmpSs programming model, which introduces a simplified pragma-based syntax for the communication between the main processor and the accelerators. Performance improvement can be achieved by the programmer explicitly specifying the data memory accesses or copies. As expected, the newer SMP+GPU system studied is more powerful than the older SMP+FPGA system. However the latter is enough to fulfill the requirements of our use case and we show that uses less energy when considering only the active power of the execution., This work is partially supported by the European Union H2020 project AXIOM (grant agreement n. 645496), HiPEAC (grant agreement n. 687698), and Mont-Blanc (grant agreements n. 288777, 610402 and 671697), the Spanish Government Programa Severo Ochoa (SEV-2015-0493), the Spanish Ministry of Science and Technology (TIN2015- 65316-P) and the Departament d’Innovació, Universitats i Empresa de la Generalitat de Catalunya, under project MPEXPAR: Models de Programaci´o i Entorns d’Execució Paral·lels (2014-SGR-1051)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Wavefront parallelization of recurrent neural networks on multi-core architectures

  • Sharma, Robin Kumar
  • Casas, Marc|||0000-0003-4564-2093
Recurrent neural networks (RNNs) are widely used for natural language processing, time-series prediction, or text analysis tasks. The internal structure of RNNs inference and training in terms of data or control dependencies across their fundamental numerical kernels complicate the exploitation of model parallelism, which is the reason why just data-parallelism has been traditionally applied to accelerate RNNs. This paper presents W-Par (Wavefront-Parallelization), a comprehensive approach for RNNs inference and training on CPUs that relies on applying model parallelism into RNNs models. We use fine-grained pipeline parallelism in terms of wavefront computations to accelerate multi-layer RNNs running on multi-core CPUs. Wavefront computations have been widely applied in many scientific computing domains like stencil kernels or dynamic programming. W-Par divides RNNs workloads across different parallel tasks by defining input and output dependencies for each RNN cell. Our experiments considering different RNNs models demonstrate that W-Par achieves up to 6.6X speed-up for RNN models inference and training in comparison to current state-of-the-art implementations on modern multi-core CPU architectures. Importantly, W-Par maximizes performance on a wide range of scenarios, including different core counts or memory hierarchy configurations, without requiring any change at the source code level., This work has been supported by the European Union's Horizon 2020 research and innovation program (MB2020 project, grant agreement 779877), by the European HiPEAC Network of Excellence, by the Spanish Ministry of Economy and Competitiveness (contract TIN2015-65316-P), and by Generalitat de Catalunya (contracts 2017-SGR-1414 and 2017-SGR-1328). M. Casas has been partially supported by the Spanish Ministry of Economy, Industry, and Competitiveness under Ramon y Cajal fellowship number RYC2017-23269., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Transparent execution of task-based parallel applications in Docker with COMP Superscalar

  • Anton, Victor
  • Ramón Cortés, Cristian|||0000-0003-4170-818X
  • Ejarque, Jorge|||0000-0003-4725-5097
  • Badia Sala, Rosa Maria|||0000-0003-2941-5499
This paper presents a framework to easily build and execute parallel applications in container-based distributed computing platforms in a user transparent way. The proposed framework is a combination of the COMP Superscalar and Docker. We have built a prototype in order to evaluate how it performs by evaluating the overhead in the building, deployment and execution phases. We have observed an important gain compared with cloud environments during the building and deployment phases. In contrast, we have detected an extra overhead during the execution, which is mainly due to the multi-host Docker networking., This work is partly supported by the Spanish Government through contracts SEV-2015-0493, TIN2015-65316-P, by the Generalitat de Catalunya under contracts 2014-SGR-1051 and 2014-SGR-1272, and by the European Union under grants 676556 (MuG Project) and 690116 (EUBra-BIGSEA Project). Results presented in this paper were obtained using the Chameleon testbed supported by the NSF., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

On the reliability of hardware event monitors in MPSoCs for critical domains

  • Barrera Herrera, Javier Enrique|||0000-0003-1469-2514
  • Kosmidis, Leonidas
  • Tabani, Hamid
  • Mezzetti, Enrico|||0000-0002-1886-2931
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Fernández, Mikel
  • Bernat Nicolau, Guillem Joan
  • Cazorla Almeida, Francisco Javier
Performance Monitoring Units (PMUs) are at the heart of most-advanced timing analysis techniques to control and bound the impact of contention in Commercial Off-The-Shelf (COTS) SoCs with shared resources (e.g. GPUs and multicore CPUs). In this paper, we report discrepancies on the values obtained from the PMU event monitors and the number of events expected based on PMU event description in the processor's official documentation. Discrepancies, which may be either due to actual errors or inaccurate specifications, make PMU readings unreliable. This is particularly problematic in consideration of the critical role played by event monitors for timing analysis in domains such as automotive and avionics. This paper proposes a systematic procedure for event monitor validation. We apply it to validate event monitors in the NVIDIA Xavier and TX2, and the Zynq UltraScale+ MPSoC. We show that, while some event monitors count as expected, this is not the case for others whose discrepancies with expected values we analyze., This work has been partially supported by the Spanish Ministry of Economy and Competitiveness (MINECO) under grant TIN2015-65316-P, the SELENE European Union’s Horizon 2020 (H2020) research and innovation programme under grant agreement No 871467, and the HiPEAC Network of Excellence. MINECO partially supported Jaume Abella under Ramon y Cajal postdoctoral fellowship (RYC-2013-14717), Enrico Mezzetti under Juan de la-Cierva-Incorporacion postdoctoral fellowship (IJCI-2016-27396), and Leonidas Kosmidis under Juan de la Cierva-Formacion postdoctoral fellowship (FJCI-2017-34095)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

FlexVC: Flexible virtual channel management in low-diameter networks

  • Fuentes, Pablo
  • Vallejo, Enrique
  • Beivide Palacio, Ramon
  • Minkenberg, Cyriel
  • Valero Cortés, Mateo|||0000-0003-2917-2482
Deadlock avoidance mechanisms for lossless lowdistance networks typically increase the order of virtual channel (VC) index with each hop. This restricts the number of buffer resources depending on the routing mechanism and limits performance due to an inefficient use. Dynamic buffer organizations increase implementation complexity and only provide small gains in this context because a significant amount of buffering needs to be allocated statically to avoid congestion. We introduce FlexVC, a simple buffer management mechanism which permits a more flexible use of VCs. It combines statically partitioned buffers, opportunistic routing and a relaxed distancebased deadlock avoidance policy. FlexVC mitigates Head-of-Line blocking and reduces up to 50% the memory requirements. Simulation results in a Dragonfly network show congestion reduction and up to 37.8% throughput improvement, outperforming more complex dynamic approaches. FlexVC merges different flows of traffic in the same buffers, which in some cases makes more difficult to identify the traffic pattern in order to support nonminimal adaptive routing. An alternative denoted FlexVCminCred improves congestion sensing for adaptive routing by tracking separately packets routed minimally and nonminimally, rising throughput up to 20.4% with 25% savings in buffer area., This work has been supported by the Spanish Government (grant SEV2015-0493 of the Severo Ochoa Program), the Spanish Ministry of Economy, Industry and Competitiveness (contracts TIN2015-65316), the Spanish Research Agency (AEI/FEDER, UE - TIN2016-76635-C2-2-R), the Spanish Ministry of Education (FPU grant FPU13/00337), the Generalitat de Catalunya (contracts 2014-SGR-1051 and 2014- SGR-1272), the European Union FP7 programme (RoMoL ERC Advanced Grant GA 321253), the European HiPEAC Network of Excellence and the European Union’s Horizon 2020 research and innovation programme (Mont-Blanc project under grant agreement No 671697)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Contention-aware performance monitoring counter support for real-time MPSoCs

  • Jalle Ibarra, Javier
  • Fernández, Mikel
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Andersson, Jan
  • Patte, Mathieu
  • Fossati, Luca
  • Zulianello, Marco
  • Cazorla Almeida, Francisco Javier
Tasks running in MPSoCs experience contention delays when accessing MPSoC’s shared resources, complicating task timing analysis and deriving execution time bounds. Understanding the Actual Contention Delay (ACD) each task suffers due to other corunning tasks, and the particular hardware shared resources in which contention occurs, is of prominent importance to increase confidence on derived execution time bounds of tasks. And, whenever those bounds are violated, ACD provides information on the reasons for overruns. Unfortunately, existing MPSoC designs considered in real-time domains offer limited hardware support to measure tasks’ ACD losing all these potential benefits. In this paper we propose the Contention Cycle Stack (CCS), a mechanism that extends performance monitoring counters to track specific events that allow estimating the ACD that each task suffers from every contending task on every hardware shared resource. We build the CCS using a set of specialized low-overhead Performance Monitoring Counters for the Cobham Gaisler GR740 (NGMP) MPSoC – used in the space domain – for which we show CCS’s benefits., The research leading to these results has received funding from the European Space Agency under contracts 4000109680, 4000110157 and NPI 4000102880, and the Ministry of Science and Technology of Spain under contract TIN-2015-65316-P. Jaume Abella has been partially supported by the Ministry of Economy and Competitiveness under Ramon y Cajal postdoctoral fellowship number RYC-2013-14717., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Validating a timing simulator for the NGMP multicore processor

  • Jalle Ibarra, Javier
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Fossati, Luca
  • Zulianello, Marco
  • Cazorla Almeida, Francisco Javier
Timing simulation is a key element in multicore systems design. It enables a fast and cost effective design space exploration, allowing to simulate new architectural improvements without requiring RTL abstraction levels. Timing simulation also allows software developers to perform early testing of the timing behavior of their software without the need of buying the actual physical board, which can be very expensive when the board uses non-COTS technology. In this paper we present the validation of a timing simulator for the NGMP multicore processor, which is a 4 core processor being developed to become the reference platform for future missions of the European Space Agency., The research leading to these results has received funding from the European Space Agency under contract NPI 4000102880 and the Ministry of Science and Technology of Spain under contract TIN-2015-65316-P. Jaume Abella has been partially supported by the Ministry of Economy and Competitiveness under Ramon y Cajal postdoctoral fellowship number RYC-2013-14717., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

The secrets of the accelerators unveiled: tracing heterogeneous executions through OMPT

  • Llort, German
  • Filgueras Izquierdo, Antonio|||0000-0001-6071-3254
  • Jiménez-González, Daniel
  • Servat, Harald
  • Teruel, Xavier
  • Mercadal, Estanislao
  • Álvarez, Carlos
  • Giménez, Judit
  • Martorell Bofill, Xavier|||0000-0002-0417-3430
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
  • Labarta Mancho, Jesús José|||0000-0002-7489-4727
Heterogeneous systems are an important trend in the future of supercomputers, yet they can be hard to program and developers still lack powerful tools to gain understanding about how well their accelerated codes perform and how to improve them. Having different types of hardware accelerators available, each with their own specific low-level APIs to program them, there is not yet a clear consensus on a standard way to retrieve information about the accelerator’s performance. To improve this scenario, OMPT is a novel performance monitoring interface that is being considered for integration into the OpenMP standard. OMPT allows analysis tools to monitor the execution of parallel OpenMP applications by providing detailed information about the activity of the runtime through a standard API. For accelerated devices, OMPT also facilitates the exchange of performance information between the runtime and the analysis tool. We implement part of the OMPT specification that refers to the use of accelerators both in the Nanos++ parallel runtime system and the Extrae tracing framework, obtaining detailed performance information about the execution of the tasks issued to the accelerated devices to later conduct insightful analysis. Our work extends previous efforts in the field to expose detailed information from the OpenMP and OmpSs runtimes, regarding the activity and performance of task-based parallel applications. In this paper, we focus on the evaluation of FPGA devices studying the performance of two common kernels in scientific algorithms: matrix multiplication and Cholesky decomposition. Furthermore, this development is seamlessly applicable for the analysis of GPGPU accelerators and Intel®Xeon PhiTM co-processors operating under the OmpSs programming model., This work was partially supported by the European Union H2020 program through the AXIOM project (grant ICT-01-2014 GA 645496) and the Mont-Blanc 2 project, by the Ministerio de Economía y Competitividad, under contracts Computación de Altas Prestaciones VII (TIN2015-65316-P); Departament d'Innovació, Universitats i Empresa de la Generalitat de Catalunya, under projects MPEXPAR: Models de Programació i Entorns d'Execució Paral·lels (2014-SGR-1051) and 2009-SGR-980; the BSC-CNS Severo Ochoa program (SEV-2011-00067); the Intel-BSC Exascale Laboratory project; and the OMPT Working Group., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Teaching HPC systems and parallel programming with small-scale clusters

  • Álvarez Martí, Lluc|||0000-0003-0506-8867
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
  • Mantovani, Filippo|||0000-0003-3559-4825
In the last decades, the continuous proliferation of High-Performance Computing (HPC) systems and data centers has augmented the demand for expert HPC system designers, administrators, and programmers. For this reason, most universities have introduced courses on HPC systems and parallel programming in their degrees. However, the laboratory assignments of these courses generally use clusters that are owned, managed and administrated by the university. This methodology has been shown effective to teach parallel programming, but using a remote cluster prevents the students from experimenting with the design, set up and administration of such systems. This paper presents a methodology and framework to teach HPC systems and parallel programming using a small-scale cluster of single-board computers. These boards are very cheap, their processors are fundamentally very similar to the ones found in HPC, and they are ready to execute Linux out of the box. So they represent a perfect laboratory playground for students experiencing how to assemble a cluster, setting it up, and configuring its system software. Also, we show that these small-scale clusters can be used as evaluation platforms for both, introductory and advanced parallel programming assignments., This work is partially supported by the Spanish Government through Programa Severo Ochoa (SEV-2015-0493), by the Spanish Ministry of Science and Technology (contracts TIN2015-65316-P and FJCI-2016-30985), by the Generalitat de Catalunya (contract 2017-SGR-1414), and by the European Union’s Horizon 2020 research and innovation programme (grant agreements 671697 and 779877)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Exploiting parallelism on GPUs and FPGAs with OmpSs

  • Bosch Pons, Jaume|||0000-0002-4040-3416
  • Filgueras Izquierdo, Antonio|||0000-0001-6071-3254
  • Vidal, Miquel|||0000-0002-1973-8289
  • Jiménez González, Daniel|||0000-0001-6064-7883
  • Álvarez Martínez, Carlos|||0000-0003-0536-5183
  • Martorell, Xavier
This paper presents the OmpSs approach to deal with heterogeneous programming on GPU and FPGA accelerators. The OmpSs programming model is based on the Mercurium compiler and the Nanos++ runtime. Applications are annotated with compiler directives specifying task-based parallelism. The Mercurium compiler transforms the code to exploit the parallelism in the SMP host cores, and also to spawn work on CUDA/OpenCL devices, and FPGA accelerators. For the CUDA/OpenCL devices, the programmer needs only to insert the annotations and provide the kernel function to be compiled by the native CUDA/OpenCL compiler. In the case of the FPGAs, OmpSs uses the High-Level Synthesis tools from FPGA vendors to generate the IP configurations for the FPGA. In this paper we present the performance obtained on the matrix multiply benchmark in the Xilinx Zynq Ultrascale+, as a result of using OmpSs on this benchmark., This work is partially supported by the European Union H2020 program through the AXIOM project (grant ICT-01-2014 GA 645496) and HiPEAC (GA 687698), by the Spanish Government through Programa Severo Ochoa (SEV-2011-0067), by the Spanish Ministerio de Economia y Competitividad under contract Computacion de Altas Prestaciones VII (TIN2015-65316-P), and the Departament d’Innovacio, Universitats i Empresa de la Generalitat de Catalunya, under project MPEXPAR: Models de Programacio i Entorns d’Execucio Paral·lels (2014-SGR-1051)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Work-in-Progress Paper: An Analysis of the Impact of Dependencies on Probabilistic Timing Analysis and Task Scheduling

  • Mezzetti, Enrico|||0000-0002-1886-2931
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Hernandez, Carles
  • Cazorla, Francisco J.
Recently there has been a renewed interest for probabilistic timing analysis (PTA) and probabilistic task scheduling (PTS). Despite the number of works in both fields, the link between them is weak: works on the latter build upon a series of assumptions on the probabilistic behavior of each task - or instances (jobs) of it - that have not been shown how to be fulfilled by PTA. This paper makes a first step towards covering this gap with emphasis on providing the right meaning of pWCET estimate as understood by both PTA and PTS. We show that the main issue related to ensuring that PTS assumptions on pWCET estimates are captured by PTA relates to the dependencies among tasks, and even jobs of a given task. Both change the scope of applicability of pWCET estimates provided by PTA and hence, their use by PTS., This work has been partially supported by the Spanish Ministry of Science and Innovation under grant TIN2015-65316-P. Jaume Abella has been partially supported by the Ministry of Economy and Competitiveness under Ramon y Cajal fellowship number RYC-2013-14717. Carles Hernández is jointly funded by the Spanish Ministry of Economy and Competitiveness and FEDER funds through grant TIN2014-60404-JIN., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

An on-board algorithm implementation on an embedded GPU: A space case study

  • Rodríguez Ferrandez, Iván
  • Kosmidis, Leonidas
  • Notebaert, Olivier
  • Cazorla Almeida, Francisco Javier
  • Steenari, David
On-board processing requirements of future space missions are constantly increasing, calling for new hardware than the traditional ones used in space. Embedded GPUs are an attractive candidate offering both high performance capabilities and low power consumption, but there are no complex industrial case studies from the space domain demonstrating these advantages. In this paper we present the GPU parallelisation of an on-board algorithm, as well as its performance on a promising embedded GPU COTS platform targeting critical systems., This work is funded by ESA under the GPU4S (GPU for Space) project (ITT AO/1-9010/17/NL/AF) and an ERC grant (No. 772773). It is also partially supported by the Spanish Ministry of Economy and Competitiveness (MINECO) under grants TIN2015-65316-P and FJCI-2017-34095 and HiPEAC., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

DROM: Enabling Efficient and Effortless Malleability for Resource Managers

  • D'Amico, Marco|||0000-0002-6195-6204
  • Garcia Gasulla, Marta|||0000-0003-3682-9905
  • López, Victor
  • Jokanovic, Ana
  • Sirvent, Raül
  • Corbalán González, Julita|||0000-0002-3926-5634
In the design of future HPC systems, research in resource management is showing an increasing interest in a more dynamic control of the available resources. It has been proven that enabling the jobs to change the number of computing resources at run time, i.e. their malleability, can significantly improve HPC system performance. However, job schedulers and applications typically do not support malleability due to the common belief that it introduces additional programming complexity and performance impact. This paper presents DROM, an interface that provides efficient malleability with no effort for program developers. The running application is enabled to adapt the number of threads to the number of assigned computing resources in a completely transparent way to the user through the integration of DROM with standard programming models, such as OpenMP/OmpSs, and MPI. We designed the APIs to be easily used by any programming model, application and job scheduler or resource manager. Our experimental results from two realistic use cases analysis, based on malleability by reducing the number of cores a job is using per node and jobs co-allocation, show the potential of DROM for improving the performance of HPC systems. In particular, the workload of two MPI+OpenMP neuro-simulators are tested, reporting improvement in system metrics, such as total run time and average response time, up to 8% and 48%, respectively., This work is partially supported by the Span- ish Government through Programa Severo Ochoa (SEV-2015-0493), by the Spanish Ministry of Science and Technology through TIN2015-65316-P project, by the Generalitat de Catalunya (contract 2017-SGR-1414) and from the European Union’s Horizon 2020 under grant agreement No 785907 (HBP SGA2), Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Reducing data movement on large shared memory systems by exploiting computation dependencies

  • Barrera, I.S.
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
  • Valero Cortés, Mateo|||0000-0003-2917-2482
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Labarta Mancho, Jesús José|||0000-0002-7489-4727
  • Casas, Marc|||0000-0003-4564-2093
Shared memory systems are becoming increasingly complex as they typically integrate several storage devices. That brings different access latencies or bandwidth rates depending on the proximity between the cores where memory accesses are issued and the storage devices containing the requested data. In this context, techniques to manage and mitigate non-uniform memory access (NUMA) effects consist in migrating threads, memory pages or both and are generally applied by the system software. We propose techniques at the runtime system level to further mitigate the impact of NUMA effects on parallel applications' performance. We leverage runtime system metadata expressed in terms of a task dependency graph, where nodes are pieces of serial code and edges are control or data dependencies between them, to efficiently reduce data transfers. Our approach, based on graph partitioning, adds negligible overhead and is able to provide performance improvements up to 1.52× and average improvements of 1.12× with respect to the best state-of-the-art approach when deployed on a 288-core shared-memory system. Our approach reduces the coherence traffic by 2.28× on average with respect to the state-of-the-art., This work has been supported by the RoMoL ERC Advanced Grant (GA 321253), by the European HiPEAC Network of Excellence, by the Spanish Ministry of Economy and Competitiveness (contract TIN2015-65316-P), by the Generalitat de Catalunya (contracts 2014-SGR-1051 and 2014-SGR-1272) and by the European Union’s Horizon 2020 research and innovation programme (grant agreements 671697 and 779877). I. Sánchez Barrera has been partially supported by the Spanish Ministry of Education, Culture and Sport under Formación del Profesorado Universitario fellowship number FPU15/03612. M. Moretó has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramón y Cajal fellowship number RYC-2016-21104., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

RICH: implementing reductions in the cache hierarchy

  • Dimic, Vladimir
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Casas, Marc|||0000-0003-4564-2093
  • Ciesko, Jan
  • Valero Cortés, Mateo|||0000-0003-2917-2482
Reductions constitute a frequent algorithmic pattern in high-performance and scientific computing. Sophisticated techniques are needed to ensure their correct and scalable concurrent execution on modern processors. Reductions on large arrays represent the most demanding case where traditional approaches are not always applicable due to low performance scalability. To address these challenges, we propose RICH, a runtime-assisted solution that relies on architectural and parallel programming model extensions. RICH updates the reduction variable directly in the cache hierarchy with the help of added in-cache functional units. Our programming model extensions fit with the most relevant parallel programming solutions for shared memory environments like OpenMP. RICH does not modify the ISA, which allows the use of algorithms with reductions from pre-compiled external libraries. Experiments show that our solution achieves the performance improvements of 11.2% on average, compared to the state-of-the-art hardware-based approaches, while it introduces 2.4% area and 3.8% power overhead., This work has been supported by the RoMoL ERC Advanced Grant (GA 321253), by the European HiPEAC Network of Excellence, by the Spanish Ministry of Economy and Competitiveness (contract TIN2015-65316-P), and by Generalitat de Catalunya (contracts 2017- SGR-1414 and 2017-SGR-1328). V. Dimić has been partially supported by the Agency for Management of University and Research Grants (AGAUR) of the Government of Catalonia under Ajuts per a la contractació de personal investigador novell fellowship number 2017 FI_B 00855. M. Moretó has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramón y Cajal fellowship number RYC-2016-21104. M. Casas has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramon y Cajal fellowship number RYC-2017-23269. This manuscript has been co-authored by National Technology & Engineering Solutions of Sandia, LLC. under Contract No. DENA0003525 with the U.S. Department of Energy/National Nuclear Security Administration, Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Resilient random modulo cache memories for probabilistically-analyzable real-time systems

  • Trilla, David|||0000-0001-9402-7703
  • Hernández, Carles
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Cazorla Almeida, Francisco Javier
Fault tolerance has often been assessed separately in safety-related real-time systems, which may lead to inefficient solutions. Recently, Measurement-Based Probabilistic Timing Analysis (MBPTA) has been proposed to estimate Worst-Case Execution Time (WCET) on high performance hardware. The intrinsic probabilistic nature of MBPTA-commpliant hardware matches perfectly with the random nature of hardware faults. Joint WCET analysis and reliability assessment has been done so far for some MBPTA-compliant designs, but not for the most promising cache design: random modulo. In this paper we perform, for the first time, an assessment of the aging-robustness of random modulo and propose new implementations preserving the key properties of random modulo, a.k.a. low critical path impact, low miss rates and MBPTA compliance, while enhancing reliability in front of aging by achieving a better – yet random – activity distribution across cache sets., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

On the tailoring of CAST-32A certification guidance to real COTS multicore architectures

  • Agirre, Irune
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Azkarate-Askasua, Mikel
  • Cazorla, Francisco J.
The use of Commercial Off-The-Shelf (COTS) multicores in real-time industry is on the rise due to multicores' potential performance increase and energy reduction. Yet, the unpredictable impact on timing of contention in shared hardware resources challenges certification. Furthermore, most safety certification standards target single-core architectures and do not provide explicit guidance for multicore processors. Recently, however, CAST-32A has been presented providing guidance for software planning, development and verification in multicores. In this paper, from a theoretical level, we provide a detailed review of CAST-32A objectives and the difficulty of reaching them under current COTS multicore design trends; at experimental level, we assess the difficulties of the application of CAST-32A to a real multicore processor, the NXP P4080., This work has been partially supported by the Spanish Ministry of Economy and Competitiveness (MINECO) under grant TIN2015-65316-P and the HiPEAC Network of Excellence. Jaume Abella has been partially supported by the MINECO under Ramon y Cajal grant RYC-2013-14717., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Modeling the impact of process variations in worst-case energy consumption estimation

  • Trilla Rodríguez, David
  • Hernández Luz, Carles
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Cazorla Almeida, Francisco Javier
The advent of autonomous power-limited systems poses a new challenge for system verification. Powerful processors needed to enable autonomous operation, are typically power-hungry, jeopardizing battery duration. Therefore, guaranteeing a given battery duration requires worst-case energy consumption (WCEC) estimation for tasks running on those systems. Unfortunately, processor energy and power can suffer significant variation across different units due to process variation (PV), i.e. variability in the electrical properties of transistors and wires due to imperfect manufacturing, which challenges existing WCEC estimation methods for applications. In this paper, we propose a statistical modeling approach to capture PV impact on applications energy and a methodology to compute their WCEC capturing PV, as required to deploy portable critical devices., This work has been partially supported by the Spanish Ministry of Economy and Competitiveness (MINECO) under grant TIN2015-65316-P and the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No. 772773). MINECO partially supported Jaume Abella under Ramon y Cajal fellowship RYC-2013-14717., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Energy-aware dynamic pricing model for cloud environments

  • Liu, Peini|||0000-0003-0058-8732
  • Bravo, Gusseppe
  • Guitart Fernández, Jordi|||0000-0003-0751-3100
Energy consumption is a critical operational cost for Cloud providers. However, as commercial providers typically use fixed pricing schemes that are oblivious about the energy costs of running virtual machines, clients are not charged according to their actual energy impact. Some works have proposed energy-aware cost models that are able to capture each client’s real energy usage. However, those models cannot be naturally used for pricing Cloud services, as the energy cost is calculated after the termination of the service, and it depends on decisions taken by the provider, such as the actual placement of the client’s virtual machines. For those reasons, a client cannot estimate in advance how much it will pay. This paper presents a pricing model for virtualized Cloud providers that dynamically derives the energy costs per allocation unit and per work unit for each time period. They account for the energy costs of the provider’s static and dynamic energy consumption by sharing out them according to the virtual resource allocation and the real resource usage of running virtual machines for the corresponding time period. Newly arrived clients during that period can use these costs as a baseline to calculate their expenses in advance as a function of the number of requested allocation and work units. Our results show that providers can get comparable revenue to traditional pricing schemes, while offering to the clients more proportional prices than fixed-price models., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

ATM: approximate task memoization in the runtime system

  • Brumar, Iulian
  • Casas, Marc|||0000-0003-4564-2093
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Valero Cortés, Mateo|||0000-0003-2917-2482
  • Sohi, Gurindar S.
Redundant computations appear during the execution of real programs. Multiple factors contribute to these unnecessary computations, such as repetitive inputs and patterns, calling functions with the same parameters or bad programming habits. Compilers minimize non useful code with static analysis. However, redundant execution might be dynamic and there are no current approaches to reduce these inefficiencies. Additionally, many algorithms can be computed with different levels of accuracy. Approximate computing exploits this fact to reduce execution time at the cost of slightly less accurate results. In this case, expert developers determine the desired tradeoff between performance and accuracy for each application. In this paper, we present Approximate Task Memoization (ATM), a novel approach in the runtime system that transparently exploits both dynamic redundancy and approximation at the task granularity of a parallel application. Memoization of previous task executions allows predicting the results of future tasks without having to execute them and without losing accuracy. To further increase performance improvements, the runtime system can memoize similar tasks, which leads to task approximate computing. By defining how to measure task similarity and correctness, we present an adaptive algorithm in the runtime system that automatically decides if task approximation is beneficial or not. When evaluated on a real 8-core processor with applications from different domains (financial analysis, stencil-computation, machine-learning and linear-algebra), ATM achieves a 1.4x average speedup when only applying memoization techniques. When adding task approximation, ATM achieves a 2.5x average speedup with an average 0.7% accuracy loss (maximum of 3.2%)., This work has been supported by the RoMoL ERC Advanced Grant (GA 321253), by the Spanish Government (grant SEV2015-0493 of the Severo Ochoa Program), by the Spanish Ministry of Science and Innovation (contracts TIN2015-65316), by Generalitat de Catalunya (contracts 2014-SGR-1051 and 2014-SGR-1272) and the European HiPEAC Network of Excellence. M. Moretó has been partially supported by the Ministry of Economy and Competitiveness under Juan de la Cierva postdoctoral fellowship number JCI-2012-15047. M. Casas is supported by the Secretary for Universities and Research of the Ministry of Economy and Knowledge of the Government of Catalonia and the Cofund programme of the Marie Curie Actions of the 7th R&D Framework Programme of the European Union (Contract 2013 BP B 00243). I. Brumar has been partially supported by the Spanish Ministry of Education, Culture and Sports under grant FPU2015/12849., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Noise inspector tool

  • Utrera Iglesias, Gladys Miriam|||0000-0002-0637-1297
  • Fornés de Juan, Jordi|||0000-0002-2262-6499
  • Labarta Mancho, Jesús José|||0000-0002-7489-4727
The operating system noise can interfere with normal execution programs. This behavior is becoming especially important when scaling parallel programs and amplified with global synchronizations. This work presents a tool to detect in a non-intrusive way the alien programs that share resources with current running applications in a multicore cluster., The authors acknowledge the support of the BSC (Barcelona Supercomputing Centre). We would like to thank the anonymous reviewers for their comments, which helped us to improve the manuscript. This work has been supported by the Spanish Ministry of Science and Innovation (contract TIN2015-65316), the Spanish Ministry of Economy, Industry and Competitiveness (HAR2014-57776-P) and the Generalitat de Catalunya (2014 SGR-1051)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Tailwind: Fast and atomic RDMA-based replication

  • Taleb, Yacine
  • Stutsman, Ryan
  • Antoniu, Gabriel
  • Cortés, Toni|||0000-0002-2537-8937
Replication is essential for fault-tolerance. However, in in-memory systems, it is a source of high overhead. Remote direct memory access (RDMA) is attractive to create redundant copies of data, since it is low-latency and has no CPU overhead at the target. However, existing approaches still result in redundant data copying and active receivers. To ensure atomic data transfers, receivers check and apply only fully received messages. Tailwind is a zero-copy recovery-log replication protocol for scale-out in-memory databases. Tailwind is the first replication protocol that eliminates all CPU-driven data copying and fully bypasses target server CPUs, thus leaving backups idle. Tailwind ensures all writes are atomic by leveraging a protocol that detects incomplete RDMA transfers. Tailwind substantially improves replication throughput and response latency compared with conventional RPC-based replication. In symmetric systems where servers both serve requests and act as replicas, Tailwind also improves normal-case throughput by freeing server CPU resources for request processing. We implemented and evaluated Tailwind on RAMCloud, a low-latency in-memory storage system. Experiments show Tailwind improves RAMCloud’s normal-case request processing throughput by 1.7×. It also cuts down writes median and 99th percentile latencies by 2x and 3x respectively., This work has been supported by the BigStorage project, funded by the European Union under the Marie SklodowskaCurie Actions (H2020-MSCA-ITN-2014-642963), by the Spanish Ministry of Science and Innovation (contract TIN2015- 65316), by Generalitat de Catalunya (contract 2014- SGR-1051). This material is based upon work supported by the National Science Foundation under Grant Nos. CNS-1566175 and CNS-1750558. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation. This work was supported in part by Facebook and VMware., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Accelerating K-mer Frequency Counting with GPU and Non-Volatile Memory

  • Cadenelli, Nicola
  • Polo Cantero, José|||0000-0002-9917-1795
  • Carrera Pérez, David|||0000-0003-4898-3424
The emergence of Next Generation Sequencing (NGS) platforms has increased the throughput of genomic sequencing and in turn the amount of data that needs to be processed, requiring highly efficient computation for its analysis. In this context, modern architectures including accelerators and non-volatile memory are essential to enable the mass exploitation of these bioinformatics workloads. This paper presents a redesign of the main component of a state-of-the-art reference-free method for variant calling, SMUFIN, which has been adapted to make the most of GPUs and NVM devices. SMUFIN relies on counting the frequency of k-mers (substrings of length k) in DNA sequences, which also constitutes a well-known problem for many bioinformatics workloads, such as genome assembly. We propose techniques to improve the efficiency of k-mer counting and to scale-up workloads like SMUFIN that used to require 16 nodes of Marenostrum 3 to a single machine with a GPU and NVM drives. Results show that although the single machine is not able to improve the time to solution of 16 nodes, its CPU time is 7.5x shorter than the aggregate CPU time of the 16 nodes, with a reduction in energy consumption of 5.5x. © 2017 IEEE., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Performance analysis and optimization of automotive GPUs

  • Mazzocchetti, Fabio
  • Benedicte Illescas, Pedro|||0000-0003-1670-7783
  • Tabani, Hamid
  • Kosmidis, Leonidas
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Cazorla Almeida, Francisco Javier
© 2019 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes,creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works., Advanced Driver Assistance Systems (ADAS) and Autonomous Driving (AD) have drastically increased the performance demands of automotive systems. Suitable highperformance platforms building upon Graphic Processing Units (GPUs) have been developed to respond to this demand, being NVIDIA Jetson TX2 a relevant representative. However, whether high-performance GPU configurations are appropriate for automotive setups remains as an open question. This paper aims at providing light on this question by modelling an automotive GPU (Jetson TX2), analyzing its microarchitectural parameters against relevant benchmarks, and identifying specific configurations able to meaningfully increase performance within similar cost envelopes, or to decrease costs preserving original performance levels. Overall, our analysis opens the door to the optimization of automotive GPUs for further system efficiency., This work has been partially supported by the Spanish Ministry of Economy and Competitiveness (MINECO) under grant TIN2015-65316-P, the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No. 772773) and the HiPEAC Network of Excellence. Pedro Benedicte and Jaume Abella have been partially supported by the MINECO under FPU15/01394 grant and Ramon y Cajal postdoctoral fellowship number RYC-2013-14717 respectively and Leonidas Kosmidis under Juan de la Cierva-Formacin postdoctoral fellowship (FJCI-2017-34095)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Low-latency multi-threaded ensemble learning for dynamic big data streams

  • Marron, Diego
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
  • Herrero Zaragoza, José Ramón|||0000-0002-4060-367X
  • Read, Jesse
  • Bifet Figuerol, Albert Carles
© 2018 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes,creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works., Real–time mining of evolving data streams involves new challenges when targeting today’s application domains such as the Internet of the Things: increasing volume, velocity and volatility requires data to be processed on–the–fly with fast reaction and adaptation to changes. This paper presents a high performance scalable design for decision trees and ensemble combinations that makes use of the vector SIMD and multicore capabilities available in modern processors to provide the required throughput and accuracy. The proposed design offers very low latency and good scalability with the number of cores on commodity hardware when compared to other state–of–the art implementations. On an Intel i7-based system, processing a single decision tree is 6x faster than MOA (Java), and 7x faster than StreamDM (C++), two well- known reference implementations. On the same system, the use of the 6 cores (and 12 hardware threads) available allow to process an ensemble of 100 learners 85x faster that MOA while providing the same accuracy. Furthermore, our solution is highly scalable: on an Intel Xeon socket with large core counts, the proposed ensemble design achieves up to 16x speed-up when employing 24 cores with respect to a single threaded execution., This work is partially supported by the Spanish Government through Programa Severo Ochoa (SEV-2015-0493), by the Spanish Ministry of Science and Technology through TIN2015-65316-P project, by the Generalitat de Catalunya (contract 2014-SGR-1051), by the Universitat Politècnica de Catalunya through an FPI/UPC scholarship and by NVIDIA through the UPC/BSC GPU Center of Excellence., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Predicting access to persistent objects through static code analysis

  • Touma, Rizkallah
  • Queralt Calafat, Anna|||0000-0003-2782-2955
  • Cortés, Toni|||0000-0002-2537-8937
  • Pérez Hernandez, María S.
In this paper, we present a fully-automatic, high-accuracy approach to predict access to persistent objects through static code analysis of object-oriented applications. The most widely-used previous technique uses a simple heuristic to make the predictions while approaches that offer higher accuracy are based on monitoring application execution. These approaches add a non-negligible overhead to the application’s execution time and/or consume a considerable amount of memory. By contrast, we demonstrate in our experimental study that our proposed approach offers better accuracy than the most common technique used to predict access to persistent objects, and makes the predictions farther in advance, without performing any analysis during application execution, This work has been supported by the European Union’s Horizon 2020 research and innovation program (grant H2020-MSCA-ITN-2014-642963), the Spanish Government (grant SEV2015-0493 of the Severo Ochoa Program), the Spanish Ministry of Science and Innovation (contract TIN2015-65316) and Generalitat de Catalunya (contract 2014-SGR-1051). The authors would also like to thank Alex Barceló for his feedback on the formalization included in this paper., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Improving HPC system throughput and response time using memory disaggregation

  • Vieira Zacarias, Felippe
  • Carpenter, Paul Matthew
  • Petrucci, Vinicius
© 2021 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes,creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works., HPC clusters are cost-effective, well understood, and scalable, but the rigid boundaries between compute nodes may lead to poor utilization of compute and memory resources. HPC jobs may vary, by orders of magnitude, in memory consumption per core. Thus, even when the system is provisioned to accommodate normal and large capacity nodes, a mismatch between the system and the memory demands of the scheduled jobs can lead to inefficient usage of both memory and compute resources. Disaggregated memory has recently been proposed as a way to mitigate this problem by flexibly allocating memory capacity across cluster nodes. This paper presents a simulation approach for at-scale evaluation of job schedulers with disaggregated memories and it introduces a new disaggregated-aware job allocation policy for the Slurm resource manager. Our results show that using disaggregated memories, depending on the imbalance between the system and the submitted jobs, a similar throughput and job response time can be achieved on a system with up to 33% less total memory provisioning., This work is part of a project that has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 754337 (EuroEXA); it has been supported by the Spanish Ministry of Science and Innovation (project TIN2015-65316-P and Ramon y Cajal fellowship RYC2018-025628-I), Generalitat de Catalunya (contracts 2014-SGR-1051 and 2014-SGR-1272), and the Severo Ochoa Programme (SEV-2015-0493)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

General purpose task-dependence management hardware for task-based dataflow programming models

  • Tan, Xubin
  • Bosch, Jaume
  • Vidal, Miquel|||0000-0002-1973-8289
  • Alvarez, Carlos
  • Jimenez-Gonzalez, Daniel
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
  • Valero Cortés, Mateo|||0000-0003-2917-2482
Task-based programming models such as OpenMP, IntelTBB and OmpSs offer the possibility of expressing dependences among tasks to drive their execution at runtime. Managing these dependences introduces noticeable overheads when targeting fine-grained tasks, diminishing the potential speedups or even introducing performance losses. To overcome this drawback, we present a general purpose hardware accelerator, Picos++, to manage the inter-task dependences efficiently in both time and energy. Our design also includes a novel nested task support. To this end, a new hardware/software co-design is presented to overcome the fact that nested tasks with dependences could result in system deadlocks due to the limited amount of resources in hardware task dependence managers. In this paper we describe a detailed implementation of this design and evaluate a parallel task-based programming model using Picos++ in a Linux embedded system with two ARM Cortex-A9 and a FPGA. The scalability and energy consumption of the real system implemented have been studied and compared against a software runtime. Even in a system limited to 2 threads, using Picos++ results in more than 1.8x speedup and 40% of energy savings in the most demanding parallelizations of real benchmarks. As a matter of fact, a hardware task dependence manager should be able to achieve much higher speedup and provide more energy savings with more threads., This work is supported by the Spanish Government (projects SEV-2015-0493 and TIN2015-65316-P), by the Generalitat de Catalunya (2014-SGR-1051 and 2014-SGR- 1272), by the European Research Council (RoMoL GA 321253) and by the “Port of OmpSs to the Android platform and Hardware support for Nanos++ runtime” Project Cooperation Agreement with LG Electronics. We also thank the Xilinx University Program., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

An Architecture for Programming Distributed Applications on Fog to Cloud Systems

  • Lordan Gomis, Francesc-Josep
  • Lezzi, Daniele|||0000-0001-5081-7244
  • Ejarque, Jorge|||0000-0003-4725-5097
  • Badia Sala, Rosa Maria|||0000-0003-2941-5499
This paper presents a framework to develop and execute applications in distributed and highly dynamic computing systems composed of cloud resources and fog devices such as mobile phones, cloudlets, and micro-clouds. The work builds on the COMPSs programming framework, which includes a programming model and a runtime already validated in HPC and cloud environments for the transparent execution of parallel applications. As part of the proposed contribution, COMPSs has been enhanced to support the execution of applications on mobile platforms that offer GPUs and CPUs. The scheduling component of COMPSs is under design to be able to offload the computation to other fog devices in the same level of the hierarchy and to cloud resources when more computational power is required. The framework has been tested executing a sample application on a mobile phone offloading task to a laptop and a private cloud., This work is partly supported by the Spanish Ministry of Science and Technology through project TIN2015-65316-P and grant BES-2013-067167, by the Generalitat de Catalunya under contracts 2014-SGR-1051 and 2014-SGR-1272, and by the European Union through the Horizon 2020 research and innovation programme under grant 730929 (mF2C Project)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Adaptive and architecture-independent task granularity for recursive applications

  • Navarro, Antoni
  • Mateo, Sergi
  • Perez, Jose M.
  • Beltran Querol, Vicenç
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
In the last few decades, modern applications have become larger and more complex. Among the users of these applications, the need to simplify the process of identifying units of work increased as well. With the approach of tasking models, this want has been satisfied. These models make scheduling units of work much more user-friendly. However, with the arrival of tasking models, came granularity management. Discovering an application’s optimal granularity is a frequent and sometimes challenging task for a wide range of recursive algorithms. Often, finding the optimal granularity will cause a substantial increase in performance. With that in mind, the quest for optimality is no easy task. Many aspects have to be considered that are directly related to lack or excess of parallelism in applications. There is no general solution as the optimal granularity depends on both algorithm and system characteristics. One commonly used method to find an optimal granularity consists in experimentally tuning an application with different granularities until an optimal is found. This paper proposes several heuristics which, combined with the appropriate monitoring techniques, allow a runtime system to automatically tune the granularity of recursive applications. The solution is independent of the architecture, execution environment or application being tested. A reference implementation in OmpSs—a task-parallel programming model—shows the programmability, ease of use and competitive performance of the proposed solution. Results show that the proposed solution is able to achieve, for any scenario, at least 75% of the performance of optimally tuned applications., This work has been supported by the Spanish Ministry of Science and Innovation (contract TIN2015-65316), the grant SEV-2015-0493 of Severo Ochoa Program awarded by the Spanish Government, and by Generalitat de Catalunya (contract 2014-SGR-1051), Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

LAEC: Look-Ahead Error Correction Codes in Embedded Processors L1 Data Cache

  • Benedicte Illescas, Pedro|||0000-0003-1670-7783
  • Hernandez, Carles
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Cazorla Almeida, Francisco Javier
As implementation technology shrinks, the presence of errors in cache memories is becoming an increasing issue in all computing domains. Critical systems, e.g. space and automotive, are specially exposed and susceptible to reliability issues. Furthermore, hardware designs in these systems are migrating to multilevel cache multicore systems, in which write-through first level data (DL1) caches have been shown to heavily harm average and guaranteed performance. While write-back DL1 caches solve this problem they come with their own challenges: they need Error Correction Codes (ECC) to tolerate soft errors, but implementing DL1 ECC in simple embedded micro-controllers requires either complex hardware to squash instructions consuming erroneous data, or delayed delivery of data to correct potential errors, which impacts performance even if such process is pipelined. In this paper we present a low-complexity hardware mechanism to anticipate data fetch and error correction in DL1 so that both (1) correct data is always delivered, but (2) avoiding additional delays in most of the cases. This achieves both high guaranteed performance and an effective solutions against errors., This work has been partially supported by the Spanish Ministry of Economy and Competitiveness (MINECO) under grant TIN2015-65316-P, the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No. 772773) and the HiPEAC Network of Excellence. Pedro Benedicte and Jaume Abella have been partially supported by the MINECO under FPU15/01394 grant and Ramon y Cajal postdoctoral fellowship number RYC-2013-14717 respectively., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Reduction to tridiagonal form for symmetric eigenproblems on asymmetric multicore processors

  • Alonso Jordá, Pedro
  • Catalán Pallarés, Sandra
  • Herrero Zaragoza, José Ramón|||0000-0002-4060-367X
  • Quintana Ortí, Enrique Salvador
  • Rodríguez Sánchez, Rafael
Asymmetric multicore processors (AMPs), as those present in ARM big.LITTLE technology, have been proposed as a means to address the end of Dennard power scaling law. The idea of these architectures is to activate only the type (and number) of cores that satisfy the quality of service requested by the application(s) in execution while delivering high energy efficiency. For dense linear algebra problems though, performance is of paramount importance, asking for an efficient use of all computational resources in the AMP. In response to this, we investigate how to exploit the asymmetric cores of an ARMv7 big.LITTLE AMP in order to attain high performance for the reduction to tridiagonal form, an essential step towards the solution of dense symmetric eigenvalue problems. The routine for this purpose in LAPACK is especially challenging, since half of its floating-point arithmetic operations (flops) are cast in terms of compute-bound kernels while the remaining half correspond to memory-bound kernels. To deal with this scenario: 1) we leverage a tuned implementation of the compute-bound kernels for AMPs; 2) we develop and parallelize new architecture-aware micro-kernels for the memory-bound kernels; 3) and we carefully adjust the type and number of cores to use at each step of the reduction procedure., The researchers from Universidad Jaume I were supported by project TIN2014-53495-R of MINECO and FEDER, and the FPU program of MECD. The researcher from Universitat Politecnica de Valencia was supported by the Generalitat Valenciana PROM- ETEOII/2014/003. The researcher from Universitat Politecnica de Catalunya was supported by projects TIN2015-65316-P from the Spanish Ministry of Education and 2014 SGR 1051 from the Generalitat de Catalunya, Dep. d´Innovacio, Universitats i Empresa., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Large-memory nodes for energy efficient high-performance computing

  • Živanovič, Darko
  • Radulović, Milan
  • Llort, German
  • Zaragoza, David
  • Strassburg, Janko
  • Carpenter, Paul Matthew
  • Radojković, Petar
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
Energy consumption is by far the most important contributor to HPC cluster operational costs, and it accounts for a significant share of the total cost of ownership. Advanced energy-saving techniques in HPC components have received significant research and development effort, but a simple measure that can dramatically reduce energy consumption is often overlooked. We show that, in capacity computing, where many small to medium-sized jobs have to be solved at the lowest cost, a practical energy-saving approach is to scale-in the application on large-memory nodes. We evaluate scaling-in; i.e. decreasing the number of application processes and compute nodes (servers) to solve a fixed-sized problem, using a set of HPC applications running in a production system. Using standard-memory nodes, we obtain average energy savings of 36%, already a huge figure. We show that the main source of these energy savings is a decrease in the node-hours (node_hours = #nodes x exe_time), which is a consequence of the more efficient use of hardware resources. Scaling-in is limited by the per-node memory capacity. We therefore consider using large-memory nodes to enable a greater degree of scaling-in. We show that the additional energy savings, of up to 52%, mean that in many cases the investment in upgrading the hardware would be recovered in a typical system lifetime of less than five years., Peer Reviewed