Pasar al contenido principal

UPC-COMPUTACION DE ALTAS PRESTACIONES VIII

PID2019-107255GB-C22

•
Nombre agencia financiadora Agencia Estatal de Investigación
Acrónimo agencia financiadora AEI
Programa Programa Estatal de I+D+i Orientado a los Retos de la Sociedad
Convocatoria Proyectos de I+D+i (Generación de Conocimiento y Retos Investigación)
Año convocatoria 2019
Plan Estatal 2017-2020 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario UNIVERSITAT POLITECNICA DE CATALUNYA

Resultados relacionados

Resultados totales (Incluyendo duplicados): 73
Encontrada(s) 1 página(s)

Dipòsit Digital de Documents de la UAB
Publicación Artículo científico (JournalArticle)

BIMSA, accelerating long sequence alignment using processing-in-memory

  • Alonso-Marín, Alejandro
  • Fernandez, Ivan
  • Aguado-Puig, Quim|||0000-0003-4871-3192
  • Gómez-Luna, Juan|||0000-0002-6514-1571
  • Marco-Sola, Santiago|||0000-0001-7951-3914
  • Mutlu, Onur
  • Moreto, Miquel|||0000-0002-9848-8758
Recent advances in sequencing technologies have stressed the critical role of sequence analysis algorithms and tools in genomics and healthcare research. In particular, sequence alignment is a fundamental building block in many sequence analysis pipelines and is frequently a performance bottleneck both in terms of execution time and memory usage. Classical sequence alignment algorithms are based on dynamic programming and often require quadratic time and memory with respect to the sequence length. As a result, classical sequence alignment algorithms fail to scale with increasing sequence lengths and quickly become memory-bound due to data-movement penalties. Processing-In-Memory (PIM) is an emerging architectural paradigm that seeks to accelerate memory-bound algorithms by bringing computation closer to the data to mitigate data-movement penalties. This work presents BIMSA (idirectional n-emory equence lignment), a PIM design and implementation for the state-of-the-art sequence alignment algorithm BiWFA (Bidirectional Wavefront Alignment), incorporating new hardware-aware optimizations for a production-ready PIM architecture (UPMEM). BIMSA supports aligning sequences up to 100K bases, exceeding the limitations of state-of-the-art PIM implementations. First, BIMSA achieves speedups up to 22.24× (11.95× on average) compared to state-of-the-art PIM-enabled implementations of sequence alignment algorithms. Second, achieves speedups up to 5.84× (2.83× on average) compared to the highest-performance multicore CPU implementation of BiWFA. Third, BIMSA exhibits linear scalability with the number of compute units in memory, enabling further performance improvements with upcoming PIM architectures equipped with more compute units and achieving speedups up to 9.56× (4.7× on average).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Cache-aware optimization of matrix multiplication and matrix factorizations on multicore processors

  • Martínez-Pérez, Héctor
  • Catalán, Sandra
  • Igual, Francisco D.
  • Herrero, José R.
  • Rodríguez-Sánchez, Rafael
  • Quintana-Ortí, Enrique S.|||0000-0002-5454-165X
[EN] This paper advocates for a careful customization of the special general matrix multiplication (GEMM) kernels that are invoked from blocked routines for several relevant matrix factorizations in LAPACK, in order to improve their performance on modern multicore processors with hierarchical cache memories. To achieve this, we leverage a refined analytical model to dynamically tune the cache configuration parameters of GEMM for these kernels, taking into account the matrix operands' dimensions, in order to improve cache occupation. In addition, toward the same goal, we accommodate a flexible development of architecture-specific micro-kernels for GEMM that allows us to select the option that, depending on the operands' dimensions, ameliorates cache utilization. Our experiments for the LU and QR factorizations on two platforms, equipped with ARM (NVIDIA Carmel) and x86 (AMD EPYC) multi-core processors, demonstrate the benefits of this approach in terms of a better cache utilization and, in general, higher performance. Moreover, they also reveal the delicate balance between optimizing for multi-threaded parallelism versus cache usage as well as the positive effects of software prefetching., This work was supported by grants PID2020-113656RB-C22, PID2019-107255GB, PID2021-126576NB-I00 and PID2021-123627OB-C52 of MCIN/AEI/10.13039/501100011033, by "ERDF A way of making Europe", and 2021-SGR-01007 of the Generalitat de Catalunya. Hector Martinez is a postdoctoral fellow supported by the Consejeria de Transformacion Economica, Industria, Conocimiento y Universidades de la Junta de Andalucia. Sandra Catalan was supported by the grant RYC2021-033973- I, funded by MCIN/AEI/10.13039/501100011033 and the European Union "NextGenerationEU"/PRTR.
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Enabling dynamic and intelligent workflows for HPC, data analytics, and AI convergence

  • Ejarque, Jorge
  • Badia, Rosa M.
  • Albertin, Loïc
  • Aloisio, Giovanni
  • Baglione, Enrico
  • Becerra, Yolanda
  • Boschert, Stefan
  • Berlin, Julian R.
  • D’Anca, Alessandro
  • Elia, Donatello
  • Exertier, François
  • Fiore, Sandro
  • Flich, José
  • Folch, Arnau
  • Gibbons, Steven J.
  • Koldunov, Nikolay
  • Lordan, Francesc
  • Lorito, Stefano
  • Løvholt, Finn
  • Macías, Jorge
  • Volpe, Manuela
The evolution of High-Performance Computing (HPC) platforms enables the design and execution of progressively larger and more complex workflow applications in these systems. The complexity comes not only from the number of elements that compose the workflows but also from the type of computations they perform. While traditional HPC workflows target simulations and modelling of physical phenomena, current needs require in addition data analytics (DA) and artificial intelligence (AI) tasks. However, the development of these workflows is hampered by the lack of proper programming models and environments that support the integration of HPC, DA, and AI, as well as the lack of tools to easily deploy and execute the workflows in HPC systems. To progress in this direction, this paper presents use cases where complex workflows are required and investigates the main issues to be addressed for the HPC/DA/AI convergence. Based on this study, the paper identifies the challenges of a new workflow platform to manage complex workflows. Finally, it proposes a development approach for such a workflow platform addressing these challenges in two directions: first, by defining a software stack that provides the functionalities to manage these complex workflows; and second, by proposing the HPC Workflow as a Service (HPCWaaS) paradigm, which leverages the software stack to facilitate the reusability of complex workflows in federated HPC infrastructures. Proposals presented in this work are subject to study and development as part of the EuroHPC eFlows4HPC project., This work has received funding from the European High-Performance Computing Joint Undertaking (JU) under grant agreement No 955558. The JU receives support from the European Union’s Horizon 2020 research and innovation programme and Spain, Germany, France, Italy, Poland, Switzerland and Norway. In Spain, it has received complementary funding from MCIN/AEI/10.13039/501100011033, Spain and the European Union NextGenerationEU/PRTR (contracts PCI2021-121957, PCI2021-121931, PCI2021-121944, and PCI2021-121927). In Germany, it has received complementary funding from the German Federal Ministry of Education and Research (contracts 16HPC016K, 6GPC016K, 16HPC017 and 16HPC018). In France, it has received financial support from Caisse des dépôts et consignations (CDC) under the action PIA ADEIP (project Calculateurs). In Italy, it has been preliminary approved for complimentary funding by Ministero dello Sviluppo Economico (MiSE) (ref. project prop. 2659). In Norway, it has received complementary funding from the Norwegian Research Council, Norway under project number 323825. In Switzerland, it has been preliminary approved for complimentary funding by the State Secretariat for Education, Research, and Innovation (SERI), Norway. In Poland, it is partially supported by the National Centre for Research and Development under decision DWM/EuroHPCJU/4/2021. The authors also acknowledge financial support by MCIN/AEI /10.13039/501100011033, Spain through the “Severo Ochoa Programme for Centres of Excellence in R&D” under Grant CEX2018-000797-S, the Spanish Government, Spain (contract PID2019-107255 GB) and by Generalitat de Catalunya, Spain (contract 2017-SGR-01414). Anna Queralt is a Serra Húnter Fellow., With funding from the Spanish government through the ‘Severo Ochoa Centre of Excellence’ accreditation (CEX2018-000797-S).
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

A closer look at referring expressions for video object segmentation

  • Bellver, Miriam
  • Ventura, Carles
  • Silberer, Carina
  • Kazakos, Ioannis
  • Torres, Jordi
  • Giró i Nieto, Xavier
The task of Language-guided Video Object Segmentation (LVOS) aims at generating binary masks for an object referred by a linguistic expression. When this expression unambiguously describes an object in the scene, it is named referring expression (RE). Our work argues that existing benchmarks used for LVOS are mainly composed of trivial cases, in which referents can be identified with simple phrases. Our analysis relies on a new categorization of the referring expressions in the DAVIS-2017 and Actor-Action datasets into trivial and non-trivial REs, where the non-trivial REs are further annotated with seven RE semantic categories. We leverage these data to analyze the performance of RefVOS, a novel neural network that obtains competitive results for the task of language-guided image segmentation and state of the art results for LVOS. Our study indicates that the major challenges for the task are related to understanding motion and static actions., Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature. This work was partially supported by the projects PID2019-107255GB-C22 and PID2020-117142GB-I00 funded by MCIN/ AEI /10.13039/501100011033 Spanish Ministry of Science, and the grant 2017-SGR-1414 of the Government of Catalonia. This work was also partially supported by the project RTI2018-095232-B-C22 funded by the Spanish Ministry of Science, Innovation and Universities.
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

OmpSs@FPGA framework for high performance FPGA computing

  • De Haro Ruiz, Juan Miguel|||0000-0002-7427-9118
  • 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 Bofill, Xavier|||0000-0002-0417-3430
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
  • Labarta Mancho, Jesús José|||0000-0002-7489-4727
This paper presents the new features of the OmpSs@FPGA framework. OmpSs is a data-flow programming model that supports task nesting and dependencies to target asynchronous parallelism and heterogeneity. OmpSs@FPGA is the extension of the programming model addressed specifically to FPGAs. OmpSs environment is built on top of Mercurium source to source compiler and Nanos++ runtime system. To address FPGA specifics Mercurium compiler implements several FPGA related features as local variable caching, wide memory accesses or accelerator replication. In addition, part of the Nanos++ runtime has been ported to hardware. Driven by the compiler this new hardware runtime adds new features to FPGA codes, such as task creation and dependence management, providing both performance increases and ease of programming. To demonstrate these new capabilities, different high performance benchmarks have been evaluated over different FPGA platforms using the OmpSs programming model. The results demonstrate that programs that use the OmpSs programming model achieve very competitive performance with low to moderate porting effort compared to other FPGA implementations., This work has received funding from EuroEXA project (European Union’s Horizon 2020 Research and Innovation Programme, under grant agreement No 754337), from Spanish Government (projects PID2019-107255GB and SEV-2015- 0493, grant BES-2016-078046), and Generalitat de Catalunya (contracts 2017-SGR-1414 and 2017-SGR-1328)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Performance characterization of containerization for HPC workloads on InfiniBand clusters: an empirical study

  • Liu, Peini|||0000-0003-0058-8732
  • Guitart Fernández, Jordi|||0000-0003-0751-3100
Containerization technology offers an appealing alternative for encapsulating and operating applications (and all their dependencies) without being constrained by the performance penalties of using Virtual Machines and, as a result, has got the interest of the High-Performance Computing (HPC) community to obtain fast, customized, portable, flexible, and reproducible deployments of their workloads. Previous work on this area has demonstrated that containerized HPC applications can exploit InfiniBand networks, but has ignored the potential of multi-container deployments which partition the processes that belong to each application into multiple containers in each host. Partitioning HPC applications has demonstrated to be useful when using virtual machines by constraining them to a single NUMA (Non-Uniform Memory Access) domain. This paper conducts a systematical study on the performance of multi-container deployments with different network fabrics and protocols, focusing especially on Infiniband networks. We analyze the impact of container granularity and its potential to exploit processor and memory affinity to improve applications’ performance. Our results show that default Singularity can achieve near bare-metal performance but does not support fine-grain multi-container deployments. Docker and Singularity-instance have similar behavior in terms of the performance of deployment schemes with different container granularity and affinity. This behavior differs for the several network fabrics and protocols, and depends as well on the application communication patterns and the message size. Moreover, deployments on Infiniband are also more impacted by the computation and memory allocation, and because of that, they can exploit the affinity better., We thank Lenovo for providing the testbed to run the experiments in this paper. This work was partially supported by Lenovo as part of Lenovo-BSC collaboration agreement, by the Spanish Government under contract PID2019-107255GB-C22, and by the Generalitat de Catalunya under contract 2017-SGR-1414 and under Grant No. 2020 FI-B 00257., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

NagareDB: A resource-efficient document-oriented time-series database

  • García Calatrava, Carlos
  • Becerra Fontal, Yolanda|||0000-0003-2357-7796
  • Cucchietti, Fernando|||0000-0002-9027-1263
  • Diví Cuesta, Carla
The recent great technological advance has led to a broad proliferation of Monitoring Infrastructures, which typically keep track of specific assets along time, ranging from factory machinery, device location, or even people. Gathering this data has become crucial for a wide number of applications, like exploration dashboards or Machine Learning techniques, such as Anomaly Detection. Time-Series Databases, designed to handle these data, grew in popularity, becoming the fastest-growing database type from 2019. In consequence, keeping track and mastering those rapidly evolving technologies became increasingly difficult. This paper introduces the holistic design approach followed for building NagareDB, a Time-Series database built on top of MongoDB—the most popular NoSQL Database, typically discouraged in the Time-Series scenario. The goal of NagareDB is to ease the access to three of the essential resources needed to building time-dependent systems: Hardware, since it is able to work in commodity machines; Software, as it is built on top of an open-source solution; and Expert Personnel, as its foundation database is considered the most popular NoSQL DB, lowering its learning curve. Concretely, NagareDB is able to outperform MongoDB recommended implementation up to 4.7 times, when retrieving data, while also offering a stream-ingestion up to 35% faster than InfluxDB, the most popular Time-Series database. Moreover, by relaxing some requirements, NagareDB is able to reduce the disk space usage up to 40%., This research was partly supported by the H2020 IoTwins project (Distributed Digital Twins for industrial SMEs: a big-data platform) funded by the EU under the call ICT-11-2018-2019, Grant Agreement No 857191, by the Spanish Ministry of Science and Innovation (contract PID2019- 107255GB) and by the Generalitat de Catalunya (contract 2017-SGR-1414)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Compute units in OpenMP: extensions for heterogeneous parallel programming

  • González Tallada, Marc|||0000-0002-3780-1106
  • Morancho Llena, Enrique|||0000-0003-2403-8145
This article evaluates the current support for heterogeneous OpenMP 5.2 applications regarding the simultaneous activation of host and device computing units (e.g., CPUs, GPUs, or FPGAs). The article identifies limitations in the current OpenMP specification and describes the design and implementation of novel OpenMP extensions and runtime support for heterogeneous parallel programming. The Compute Unit (CUs) abstraction is introduced in the OpenMP programming model. The Compute Unit abstraction is defined in terms of an aggregation of computing elements (e.g., CPUs, GPUs, FPGAs). On top of CUs, the article describes dynamic work sharing constructs and schedulers that address the inherent differences in compute power of host and device CUs. New constructs and the corresponding runtime support are described for the new abstractions. The article evaluates the case of a hybrid multilevel parallelization of the NPB-MZ benchmark suite. The implementation exploits both coarse-grain and fine-grain parallelism, mapped to CUs of different nature (GPUs and CPUs). All CUs are activated using the new extensions and runtime support. We compare hybrid and nonhybrid executions under two state-of-the-art work-distribution schemes (Static and Dynamic Task schedulers). On a computing node composed of one AMD EPYC 7742 @ 2.250GHz (64 cores and 2 threads/core, totalling 128 threads per node) and 2 GPU AMD Radeon Instinct MI50 with 32GB, hybrid executions present speedups from 1.08 up to 3.18 with respect to a nonhybrid GPU implementation, depending on the number of activated CUs., This work was supported by the Spanish Ministry of Science and Technology (PID2019-107255GB)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

A closer look at referring expressions for video object segmentation

  • Bellver Bueno, Míriam
  • Ventura Royo, Carles|||0000-0002-3055-339X
  • Silberer, Carina
  • Kazakos, Ioannis
  • Torres Viñals, Jordi|||0000-0003-1963-7418
  • Giró Nieto, Xavier|||0000-0002-9935-5332
The task of Language-guided Video Object Segmentation (LVOS) aims at generating binary masks for an object referred by a linguistic expression. When this expression unambiguously describes an object in the scene, it is named referring expression (RE). Our work argues that existing benchmarks used for LVOS are mainly composed of trivial cases, in which referents can be identified with simple phrases. Our analysis relies on a new categorization of the referring expressions in the DAVIS-2017 and Actor-Action datasets into trivial and non-trivial REs, where the non-trivial REs are further annotated with seven RE semantic categories. We leverage these data to analyze the performance of RefVOS, a novel neural network that obtains competitive results for the task of language-guided image segmentation and state of the art results for LVOS. Our study indicates that the major challenges for the task are related to understanding motion and static actions., Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature. This work was partially supported by the projects PID2019-107255GB-C22 and PID2020-117142GB-I00 funded by MCIN/ AEI /10.13039/501100011033 Spanish Ministry of Science, and the grant 2017-SGR-1414 of the Government of Catalonia. This work was also partially supported by the project RTI2018-095232-B-C22 funded by the Spanish Ministry of Science, Innovation and Universities., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Boosting HPC data analysis performance with the ParSoDA-Py library

  • Belcastro, Loris
  • Giampà, Salvatore
  • Marozzo, Fabrizio
  • Talia, Domenico
  • Trunfio, Paolo
  • Badia Sala, Rosa Maria|||0000-0003-2941-5499
  • Ejarque, Jorge|||0000-0003-4725-5097
  • Mammadli, Nihad
Developing and executing large-scale data analysis applications in parallel and distributed environments can be a complex and time-consuming task. Developers often find themselves diverted from their application logic to handle technical details about the underlying runtime and related issues. To simplify this process, ParSoDA, a Java library, has been proposed to facilitate the development of parallel data mining applications executed on HPC systems. It simplifies the process by providing built-in scalability mechanisms relying on the Hadoop and Spark frameworks. This paper presents ParSoDA-Py, the Python version of the ParSoDA library, which allows for further support of commonly used runtimes and libraries for big data analysis. After a complete library redesign, ParSoDA can be now easily integrated with other Python-based distributed runtimes for HPC systems, such as COMPSs and Apache Spark, and with the large ecosystem of Python-based data processing libraries. The paper discusses the adaptation process, which takes into consideration the new technical requirements, and evaluates both usability and scalability through some case study applications., This work has been partially funded by the European Commission’s Horizon 2020 Framework program and the European High-Performance Computing Joint Undertaking (JU) under Grant agreement No 955558 and by MCIN/AEI/10.13039/501100011033 and the European Union NextGenerationEU/PRTR (PCI2021-121957), project eFlows4HPC. It has also been supported by the Spanish Government (PID2019-107255GB) and by the Departament de Recerca i Universitats de la Generalitat de Catalunya to the Research Group MPiEDist (2021 SGR 00412) We also acknowledge financial support from “National Centre for HPC, Big Data and Quantum Computing," CN00000013 - CUP H23C22000360005, and from “FAIR - Future Artificial Intelligence Research" Project - CUP H23C22000860006., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Low-cost training of image-to-image diffusion models with incremental learning and task/domain adaptation

  • Antona Pizà, Héctor
  • Otero Calviño, Beatriz|||0000-0002-9194-559X
  • Tous Liesa, Rubén|||0000-0002-1409-5843
Diffusion models specialized in image-to-image translation tasks, like inpainting and colorization, have outperformed the state of the art, yet their computational requirements are exceptionally demanding. This study analyzes different strategies to train image-to-image diffusion models in a low-resource setting. The studied strategies include incremental learning and task/domain transfer learning. First, a base model for human face inpainting is trained from scratch with an incremental learning strategy. The resulting model achieves an FID score almost equivalent to that of its batch learning equivalent while significantly reducing the training time. Second, the base model is fine-tuned to perform a different task, image colorization, and, in a different domain, landscape images. The resulting colorization models showcase exceptional performances with a minimal number of training epochs. We examine the impact of different configurations and provide insights into the ability of image-to-image diffusion models for transfer learning across tasks and domains., This work is partially supported by the Spanish Ministry of Science and Innovation under contracts PID2019-107255GB and PID2021-124463OB-IOO, by the Generalitat de Catalunya under grants 2017-SGR-962, 2021-SGR-00326, 2021-SGR-00478 and DRAC (IU16-011591). The research leading to these results has received funding from the European Union’s Horizon 2020 research and innovation program under the HORIZON-AG PHOENI2X (101070586) and HORIZON-EU VITAMIN-V (101093062) projects., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Near-optimal replacement policies for shared caches in multicore processors

  • Díaz Maag, Javier
  • Ibáñez Marín, Pablo
  • Monreal Arnal, Teresa|||0000-0002-0458-2234
  • Viñals Yúfera, Victor
  • Llaberia Griñó, José M.|||0000-0002-3753-4108
An optimal replacement policy that minimizes the miss rate in a private cache was proposed several decades ago. It requires knowing the future access sequence the cache will receive. There is no equivalent for shared caches because replacement decisions alter this future sequence. We present a novel near-optimal policy for minimizing the miss rate in a shared cache that approaches the optimal execution iteratively. During each iteration, the future access sequence is reconstructed on every miss interleaving the future per-core sequences, taken from the previous iteration. This single sequence feeds a classical private-cache optimum replacement policy. Our evaluation on a shared last-level cache shows that our proposal iteratively converges to a near-optimal miss rate that is independent of the initial conditions, within a margin of 0.1%. The best state-of-the-art online policies achieve around 65% of the miss rate reduction obtained by our near-optimal proposal. In a shared cache, miss rate optimization does not imply the optimization of other metrics. Therefore, we also propose a new near-optimal policy to maximize fairness between cores. The best state-of-the-art online policy achieves 60% of the improvement in fairness seen with our near-optimal policy. Our proposals are useful both for setting upper performance bounds and inspiring implementable mechanisms for shared caches., The authors acknowledge support from grants (1) PID2019-105660RB-C21 and PID2019-107255GB-C22 from Agencia Estatal de Investigación (AEI) from Spain and European Regional Development Fund (ERDF); (2) gaZ: T58_20R research group from Aragón Government and European Social Fund (ESF); and (3) 2014-2020 "Construyendo Europa desde Aragón" from European Regional Development Fund (ERDF)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Embedded scaffolding for teaching and assessing inquiry-based hands-on laboratory on distributed systems

  • Guitart Fernández, Jordi|||0000-0003-0751-3100
Context Information Technology education must cultivate proficiency on distributed systems, including strong hands-on laboratory skills, to meet the needs of the society and the industry. Given the complexity of distributed systems, any successful methodology to teach them to novice students must be scaffolded appropriately to ensure that the students acquire the required degree of expertise. Objective We propose a comprehensive scaffolding approach for inquiry-based hands-on laboratory on a distributed systems course, which guides not only the learning process, but also its assessment. The approach is based mainly on embedded scaffolds, namely explicit coding and experimental milestones and open questions with predefined grades, but also features contingent scaffolds provided by the teacher when additional assistance is needed. Method We apply the methodology in the context of the subject ‘Distributed Network Systems’ offered by our university. We compare the students' performance during three academic courses using the proposed methodology with respect to the three previous courses that were still using the former methodology. We use both visual representations and planned Analysis of Variance (ANOVA) tests to verify our hypothesis defined as a complex contrast. Findings We find that there is a statistically significant improvement in the students' performance when using the new methodology, both in their grades of the assignments (F(1, 75.364) = 17.770, p = 6.85*10^-5) and, more importantly, also in their grades of the exam questions about the practicals (F(1, 123.186) = 13.285, p = 3.93*10^-4). Implications Our results encourage other instructors to incorporate embedded scaffolds for teaching and assessing their hands-on laboratories on distributed systems., This research was partially supported by the Agencia Estatal de Investigación under contracts PID2019-107255GB-C22 and PID2021-126248OB-I00 and by the Generalitat de Catalunya under contract 2021-SGR-00478., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Optimizing MPI collective operations for intra-node shared memory exploitation and oversubscription

  • Utrera Iglesias, Gladys Miriam|||0000-0002-0637-1297
  • Bull, J. Mark
Node sizes in multicore clusters are becoming larger, so applications should exploit the shared memory inside a node, to potentially reduce communication latencies compared to network communications. The Message Passing Interface library (MPI) serves as the de facto standard for parallel applications on distributed memory environments. The appearance of the MPI-3 shared memory extension has made the idea of optimizing the MPI collective operations at the intra-node level both attractive and portable. Taking advantage of this facility, we present a hierarchical design of the algorithms for MPI_Allreduce and MPI_Bcast collective operations, which we name Fullpar. The proposal is based on partitioning the messages and exploiting concurrency between network communication and shared-memory operations at intra-node level in the message dissemination process. Previously proposed hierarchical collective algorithms do not exploit all the available parallelism and/or the direct implementation of shared-memory at intra-node level. Furthermore, we explore the application of oversubscription, to enhance resource utilization. Oversubscribing CPUs offers opportunities to optimize resource allocation, as well as making CPUs available to other applications. We implement our proposal on top of the Intel MPI and OpenMPI libraries using the MPI profiling (PMPI) mechanism, and carry out evaluations on platforms with different architectures characteristics such as the node size and interconnection network. Experimental results show that the proposed approach often achieves lower execution times than the evaluated alternatives for medium and large message sizes. Furthermore, the introduction of oversubscription is shown to have almost no overhead, especially for the broadcast of large messages., This work was supported by the Spanish Ministry of Education (PID2019-107255GB-C22); the Generalitat de Catalunya (2021-SGR-01007); and Project HPC-EUROPA3 (INFRAIA-2016-1-730897), with the support of the EC Research Innovation Action under the H2020 Programme., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Human pose completion in partial body camera shots

  • Tous Liesa, Rubén|||0000-0002-1409-5843
  • Nin Guerrero, Jordi|||0000-0002-9659-2762
  • Igual Muñoz, Laura
Many actual images contain partial body camera shots, in which a significant part of the body is not visible. This issue is especially prevalent in film images, where less than 10% are full-body shots. Most 2D human pose estimation methods return incomplete poses when applied to partial body images. This lack of completeness becomes a problem in some situations, for example, when the 2D pose is converted to a 3D pose by a two-stage 3D human pose estimation method since most of these methods require a complete pose to work. This article proposes a new technique, called CompletePose, consisting of completing the missing keypoints when 2D human pose estimation methods are applied to images with partial body camera shots. A Conditional Generative Adversarial Network is used to obtain a complete and plausible pose, realistic enough to predict a 3D pose with a two-stage 3D human pose estimation method. A complete empirical validation has been carried out with the Human3.6 M dataset and a new dataset, called CHARADE, specially built and made public for reproducibility and benchmarking for this research. Quantitative evaluation employing the Fréchet distance shows that the approach manages to approximate the actual data distribution. The qualitative evaluation shows that the completed poses enable obtaining plausible 3D poses from images previously intractable., This work is partially supported by the Spanish Ministry of Science and Innovation under contracts PID2019-107255GB, PLEC2021-007850 and RTI2018-095232-B-C21, and by the SGR programme of the Catalan Government under contracts 2021-SGR00478, 2017-SGR-1414 and 2017-SGR-1742., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Programming parallel dense matrix factorizations and inversion for new-generation NUMA architectures

  • Catalán Pallarés, Sandra
  • Igual Peña, Francisco D.
  • Herrero Zaragoza, José Ramón|||0000-0002-4060-367X
  • Rodríguez Sánchez, Rafael
  • Quintana Ortí, Enrique Salvador
We propose a methodology to address the programmability issues derived from the emergence of new-generation shared-memory NUMA architectures. For this purpose, we employ dense matrix factorizations and matrix inversion (DMFI) as a use case, and we target two modern architectures (AMD Rome and Huawei Kunpeng 920) that exhibit configurable NUMA topologies. Our methodology pursues performance portability across different NUMA configurations by proposing multi-domain implementations for DMFI plus a hybrid task- and loop-level parallelization that configures multi-threaded executions to fix core-to-data binding, exploiting locality at the expense of minor code modifications. In addition, we introduce a generalization of the multi-domain implementations for DMFI that offers support for virtually any NUMA topology in present and future architectures. Our experimentation on the two target architectures for three representative dense linear algebra operations validates the proposal, reveals insights on the necessity of adapting both the codes and their execution to improve data access locality, and reports performance across architectures and inter- and intra-socket NUMA configurations competitive with state-of-the-art message-passing implementations, maintaining the ease of development usually associated with shared-memory programming., This research was sponsored by project PID2019-107255GB of Ministerio de Ciencia, Innovación y Universidades; project S2018/TCS-4423 of Comunidad de Madrid; project 2017-SGR-1414 of the Generalitat de Catalunya and the Madrid Government under the Multiannual Agreement with UCM in the line Program to Stimulate Research for Young Doctors in the context of the V PRICIT, project PR65/19-22445. This project has also received funding from the European High-Performance Computing Joint Undertaking (JU) under grant agreement No 955558. The JU receives support from the European Union’s Horizon 2020 research and innovation programme, and Spain, Germany, France, Italy, Poland, Switzerland, Norway. The work is also supported by grants PID2020-113656RB-C22 and PID2021-126576NB-I00 of MCIN/AEI/10.13039/501100011033 and by ERDF A way of making Europe., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Mixed-precision pre-pivoting strategy for the LU factorization

  • Sahraneshinsamani, Nima
  • Catalán Pallarés, Sandra
  • Herrero Zaragoza, José Ramón|||0000-0002-4060-367X
This paper investigates the efficient application of half-precision floating-point (FP16) arithmetic on GPUs for boosting LU decompositions in double (FP64) precision. Addressing the motivation to enhance computational efficiency, we introduce two novel algorithms: Pre-Pivoted LU (PRP) and Mixed-precision Panel Factorization (MPF). Deployed in both hybrid CPU-GPU setups and native GPU-only configurations, PRP identifies pivot lists through LU decomposition computed in reduced precision and subsequently reorders matrix rows in FP64 precision before executing LU decomposition without pivoting. Two variants of PRP, namely hPRP and xPRP, are introduced, differing in their computation of pivot lists in full half-precision or mixed half-single precision. The MPF algorithm generates FP64 LU factorization while internally utilizing hPRP for panel factorization, showcasing accuracy on par with standard DGETRF but with superior speed. The study further explores auxiliary functions required for the native mode implementation of PRP variants and MPF., The authors acknowledge the funding received from project PID2019-107255GB of the Spanish Ministerio de Economía Competitividad, and project 2021-SGR-01007 from the Generalitat de Catalunya, Spain. Sandra Catalán was supported by the grant RYC2021-033973-I, funded by MCIN/AEI/10.13039/501100011033 and the European Union “NextGenerationEU”/PRTR, and UJI-2023-04, funded by Universitat Jaume I., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Evaluation of work distribution schedulers for heterogeneous architectures and scientific applications

  • González Tallada, Marc|||0000-0002-3780-1106
  • Morancho Llena, Enrique|||0000-0003-2403-8145
This article explores and evaluates variants of state-of-the-art work distribution schemes adapted for scientific applications running on hybrid systems. A hybrid implementation (multi-GPU and multi-CPU) of the NASA Parallel Benchmarks - MutiZone (NPB-MZ) benchmarks is described to study the different elements that condition the execution of this suite of applications when parallelism is spread over a set of computing units (CUs) of different computational power (e.g., GPUs and CPUs). This article studies the influence of the work distribution schemes on the data placement across the devices and the host, which in turn determine the communications between the CUs, and evaluates how the schedulers are affected by the relationship between data placement and communications. We show that only the schedulers aware of the different computational power of the CUs and minimize communications are able to achieve an appropriate work balance and high performance levels. Only then does the combination of GPUs and CPUs result in an effective parallel implementation that boosts the performance of a non-hybrid multi-GPU implementation. The article describes and evaluates the schedulers static-pcf, Guided, and Clustered Guided to solve the previously mentioned limitations that appear in hybrid systems. We compare them against state-of-the-art static and memorizing dynamic schedulers. Finally, on a system with an AMD EPYC 7742 at 2.250GHz (64 cores, 2 threads per core, 128 threads) and two AMD Radeon Instinct MI50 GPUs with 32GB, we have observed that hybrid executions speed up from 1.1× up to 3.5× with respect to a non-hybrid GPU implementation., This work was supported by the Spanish Ministry of Science and Technology under the grant no. PID2019-107255GB., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

HRM: merging hardware event monitors for improved timing analysis of complex MPSoCs

  • Vilardell Moreno, Sergi
  • Serra Mochales, Isabel
  • Santalla, Roberto
  • Mezzetti, Enrico|||0000-0002-1886-2931
  • Abella Ferrer, Jaume|||0000-0001-7951-4028
  • Cazorla Almeida, Francisco Javier
The Performance Monitoring Unit (PMU) in MPSoCs is at the heart of the latest measurement-based timing analysis techniques in Critical Embedded Systems. In particular, hardware event monitors (HEMs) in the PMU are used as building blocks in the process of budgeting and verifying software timing by tracking and controlling access counts to shared resources. While the number of HEMs in current MPSoCs reaches hundreds, they are read via Performance Monitoring Counters whose number is usually limited to 4-8, thus requiring multiple runs of each experiment in order to collect all desired HEMs. Despite the effort of engineers in controlling the execution conditions of each experiment, the complexity of current MPSoCs makes it arguably impossible to completely remove the noise affecting each run. As a result, HEMs read in different runs are subject to different variability, and hence, those HEMs captured in different runs cannot be ‘blindly’ merged. In this work, we focus on the NXP T2080 platform where we observed up to 59% variability across different runs of the same experiment for some relevant HEMs (e.g. processor cycles). We develop a HEM reading and merging (HRM) approach to join reliably HEMs across different runs as a fundamental element of any measurement-based timing budgeting and verification technique. Our method builds on order statistics and the selection of an anchor HEM read in all runs to derive the most plausible combination of HEM readings that keep the distribution of each HEM and their relationship with the anchor HEM intact., This work has been partially supported by the Spanish Ministry of Science and Innovation under grant PID2019-107255GB, 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., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Performance comparison of multi-container deployment schemes for HPC workloads: an empirical study

  • Liu, Peini|||0000-0003-0058-8732
  • Guitart Fernández, Jordi|||0000-0003-0751-3100
The high-performance computing (HPC) community has recently started to use containerization to obtain fast, customized, portable, flexible, and reproducible deployments of their workloads. Previous work showed that deploying an HPC workload into a single container can keep bare-metal performance. However, there is a lack of research on multi-container deployments that partition the processes belonging to each application into different containers. Partitioning HPC applications has shown to improve their performance on virtual machines by allowing to set affinity to a non-uniform memory access (NUMA) domain for each of them. Consequently, it is essential to understand the performance implications of distinct multi-container deployment schemes for HPC workloads, focusing on the impact of the container granularity and its combination with processor and memory affinity. This paper presents a systematic performance comparison and analysis of multi-container deployment schemes for HPC workloads on a single-node platform, which considers different containerization technologies (including Docker and Singularity), two different platform architectures (UMA and NUMA), and two application subscription modes (exact subscription and over-subscription). Our results indicate that finer-grained multi-container deployments, on the one side, can benefit the performance of some applications with low interprocess communication, especially in over-subscribed scenarios and when combined with affinity, but, on the other side, they can incur some performance degradation for communication-intensive applications when using containerization technologies that deploy isolated network namespaces., This work was partially supported by Lenovo as part of Lenovo-BSC collaboration agreement, by the Spanish Government under contract PID2019-107255GB-C22, and by the Generalitat de Catalunya under contract 2017-SGR-1414 and under grant 2020 FI-B 00257., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Automatic distributed deep learning using resource-constrained edge devices

  • Gutiérrez Torre, Alberto|||0000-0002-5548-3359
  • Bahadori, Kiyana
  • Baig, Shuja-ur-Rehman
  • Iqbal, Waheed
  • Vardanega, Tullio
  • Berral García, Josep Lluís|||0000-0003-3037-3580
  • Carrera Pérez, David|||0000-0003-4898-3424
Processing data generated at high volume and speed from the Internet of Things, smart cities, domotic, intelligent surveillance, and e-healthcare systems require efficient data processing and analytics services at the Edge to reduce the latency and response time of the applications. The Fog Computing Edge infrastructure consists of devices with limited computing, memory, and bandwidth resources, which challenge the construction of predictive analytics solutions that require resource-intensive tasks for training machine learning models. In this work, we focus on the development of predictive analytics for urban traffic. Our solution is based on deep learning techniques localized in the Edge, where computing devices have very limited computational resources. We present an innovative method for efficiently training of Gated Recurrent-Units (GRUs) across available resource-constrained CPU and GPU Edge devices. Our solution employs distributed GRU model learning and dynamically stops the training process to utilize the low-power and resource-constrained Edge devices while ensuring good estimation accuracy effectively. The proposed solution was extensively evaluated using low-powered ARM-based devices, including Raspberry Pi v3 and the low-powered GPU-enabled device NVIDIA Jetson Nano, and also compared them with Single-CPU Intel Xeon machines. For the evaluation experiments, we used real-world Floating Car Data. The experiments show that the proposed solution delivers excellent prediction accuracy and computational performance on the Edge when compared with the baseline methods., This work has been partially supported by the Spanish Government (contract PID2019-107255GB), the Generalitat de Catalunya (contract 2014-SGR-1051), the University of Padova, and Severo Ochoa CoE (SEV-2015-0493-16-5). The statements made herein are solely the responsibility of the authors., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Pictonaut: movie cartoonization using 3D human pose estimation and GANs

  • Tous Liesa, Rubén|||0000-0002-1409-5843
This article describes Pictonaut, a novel method to automatically synthetise animated shots from motion picture footage. Its results are editable (backgrounds, characters, lighting, etc.) with conventional 3D software, and they have the finish of professional 2D animation. Rather than addressing the challenge solely as an image translation problem, a hybrid approach combining multi-person 3D human pose estimation and GANs is taken. Sub-sampled video frames are processed with OpenPose and SMPLify-X to obtain the 3D parameters of the pose (body, hands and face expression) of all depicted characters. The captured parameters are retargeted into manually selected 3D models, cel shaded to mimic the style of a 2D cartoon. The results of sub-sampled frames are interpolated to generate a complete and smooth motion for all the characters. The background is cartoonized with a GAN. Qualitative evaluation shows that the approach is feasible, and a small dataset of synthetised shots obtained from real movie scenes is provided., This work is partially supported by the Spanish Ministry of Science and Innovation under contract PID2019-107255GB, and by the SGR programme 2017-SGR-1414 of the Catalan Government., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Adaptive power shifting for power-constrained heterogeneous systems

  • Ortega Carrasco, Cristobal
  • Alvarez Martí, Lluc
  • Buyuktosunoglu, Alper
  • Bertran Monfort, Ramon
  • Rosedahl, Todd Jon
  • Bose, Pradip
  • Moretó Planas, Miquel|||0000-0002-9848-8758
The number and heterogeneity of compute devices, even within a single compute node, has been steadily on the rise. Since all systems must operate under a power cap, the number of discrete devices that can run simultaneously at their highest frequency is limited by the globally-imposed power cap. Current systems incorporate a centralized power management unit that statically controls the distribution of power among the devices within the node. However, such static distribution policies are unaware of the dynamic utilization profile across the devices, which leads to unfair power allocations that end up degrading system throughput performance. The problem is particularly acute in the presence of heterogeneity since type-specific performance-boost capabilities cannot be leveraged via utilization-agnostic static power allocations. This paper proposes Adaptive Power Shifting for multi-accelerator heterogeneous systems (APS), a technique that leverages system utilization information to dynamically allocate and re-distribute power budgets across multiple discrete devices. Democratizing the power allocation based on dynamic needs results in dramatic speedup over a need-agnostic static allocation. We use APS in a real OpenPOWER compute node with 2 CPUs and 4 GPUs to demonstrate the value of on-demand, equitable power allocations. Overall, the proposed solution increases performance with respect to two state-of-the-art techniques by up to 14.9% and 13.8%., This work has been partially supported by the European Union’s Horizon 2020 research and innovation program under the Mont-Blanc 2020 project (grant agreement 779877), by the Spanish Ministry of Science and Innovation (contract PID2019-107255GB-C22), by Generalitat de Catalunya (contracts 2014-SGR-1051 and 2014-SGR-1272) and by the IBM/BSC Deep Learning Center initiative. Ll. Alvarez has been supported in part by the Spanish Ministry of Economy, Industry and Competitiveness under the Juan de la Cierva Formacion fellowship No. FJCI-2016- 30984. M. Moreto has been supported in part by the Spanish Ministry of Economy, Industry and Competitiveness under Ramon y Cajal fellowship No. RYC-2016-21104., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Temporally coherent video cartoonization for animation scenery generation

  • Rayo Hernández, Gustavo Enrique
  • Tous Liesa, Rubén|||0000-0002-1409-5843
The automatic transformation of short background videos from real scenarios into other forms with a visually pleasing style, like those used in cartoons, holds application in various domains. These include animated films, video games, advertisements, and many other areas that involve visual content creation. A method or tool that can perform this task would inspire, facilitate, and streamline the work of artists and people who produce this type of content. This work proposes a method that integrates multiple components to translate short background videos into other forms that contain a particular style. We apply a fine-tuned latent diffusion model with an image-to-image setting, conditioned with the image edges (computed with holistically nested edge detection) and CLIP-generated prompts to translate the keyframes from a source video, ensuring content preservation. To maintain temporal coherence, the keyframes are translated into grids and the style is interpolated with an example-based style propagation algorithm. We quantitatively assess the content preservation and temporal coherence using CLIP-based metrics over a new dataset of 20 videos translated into three distinct styles., This work is partially supported by the Spanish Ministry of Science and Innovation under contract PID2019-107255GB, and by the SGR programme 2021-SGR-00478 of the Catalan Government., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Cache-aware optimization of matrix multiplication and matrix factorizations on multicore processors

  • Martínez Pérez, Héctor
  • Catalán Pallarés, Sandra
  • Igual Peña, Francisco D.
  • Herrero Zaragoza, José Ramón|||0000-0002-4060-367X
  • Rodríguez Sánchez, Rafael
  • Quintana Ortí, Enrique Salvador
This paper advocates for a careful customization of the special general matrix multiplication (GEMM) kernels that are invoked from blocked routines for several relevant matrix factorizations in LAPACK, in order to improve their performance on modern multicore processors with hierarchical cache memories. To achieve this, we leverage a refined analytical model to dynamically tune the cache configuration parameters of GEMM for these kernels, taking into account the matrix operands’ dimensions, in order to improve cache occupation. In addition, toward the same goal, we accommodate a flexible development of architecture-specific micro-kernels for GEMM that allows us to select the option that, depending on the operands’ dimensions, ameliorates cache utilization. Our experiments for the LU and QR factorizations on two platforms, equipped with ARM (NVIDIA Carmel) and x86 (AMD EPYC) multi-core processors, demonstrate the benefits of this approach in terms of a better cache utilization and, in general, higher performance. Moreover, they also reveal the delicate balance between optimizing for multi-threaded parallelism versus cache usage as well as the positive effects of software prefetching., This work was supported by grants PID2020- 113656RB-C22, PID2019-107255GB, PID2021-126576NB-I00 and PID2021-123627OB-C52 of MCIN/AEI/10.13039/501100011033, by ‘‘ERDF A way of making Europe’’, and 2021-SGR-01007 of the Generalitat de Catalunya. Héctor Martínez is a postdoctoral fellow supported by the Consejería de Transformación Económica, Industria, Conocimiento y Universidades de la Junta de Andalucía. Sandra Catalán was supported by the grant RYC2021-033973- I, funded by MCIN/AEI/10.13039/501100011033 and the European Union ‘‘NextGenerationEU’’/PRTR. Funding Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Open project-based learning for dynamic adaptability of IT education

  • Tous Liesa, Rubén|||0000-0002-1409-5843
  • Freitag, Fèlix|||0000-0001-5438-479X
The pace of technological change is accelerating, the duration of the trends is becoming shorter and the number of technological alternatives to solve IT problems grows quickly. Keeping IT higher education course curricula up-to-date becomes more and more challenging. This paper presents an empirical study that aims to assess whether an open-statement and open-solution methodology in an IT-related project-based learning (PBL) course led to the adaptation of the skills learned by students to the trends of the main IT technology domains. The study analyses data from more than 90 projects by students from eight academic years of the course” Project on Information Technologies”, carried out by students in the last year of the Bachelor of Science in Informatics Engineering at the Barcelona School of Informatics. Our results identify a high correlation between the technologies chosen in student projects and technological trends, which shows that the methodology encourages students to incorporate the latest technological innovations in their project work plan., This work is partially supported by the Spanish Ministry of Economy and Competitivity under contract TIN2015-65316-P, by the Spanish Ministry of Science and Innovation under contracts PID2019-107255GB and PID2019-106774RB-C21, and by the SGR programmes (2017-SGR-962 and 2017-SGR-990) of the Catalan Government., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

SpChar: characterizing the sparse puzzle via decision trees

  • Sgherzi, Francesco
  • Siracusa, Marco
  • Fernández Vega, Ivan
  • Armejach Sanosa, Adrià|||0000-0003-2869-668X
  • Moretó Planas, Miquel|||0000-0002-9848-8758
Sparse matrix computation is crucial in various modern applications, including large-scale graph analytics, deep learning, and recommender systems. The performance of sparse kernels varies greatly depending on the structure of the input matrix, making it difficult to gain a comprehensive understanding of sparse computation and its relationship to inputs, algorithms, and target machine architecture. Despite extensive research on certain sparse kernels, such as Sparse Matrix-Vector Multiplication (SpMV), the overall family of sparse algorithms has yet to be investigated as a whole. This paper introduces SpChar, a workload characterization methodology for general sparse computation. SpChar employs tree-based models to identify the most relevant hardware and input characteristics, starting from hardware and input-related metrics gathered from Performance Monitoring Counters (PMCs) and matrices. Our analysis enables the creation of a characterization loop that facilitates the optimization of sparse computation by mapping the impact of architectural features to inputs and algorithmic choices. We apply SpChar to more than 600 matrices from the SuiteSparse Matrix collection and three state-of-the-art Arm Central Processing Units (CPUs) to determine the critical hardware and software characteristics that affect sparse computation. In our analysis, we determine that the biggest limiting factors for high-performance sparse computation are (1) the latency of the memory system, (2) the pipeline flush overhead resulting from branch misprediction, and (3) the poor reuse of cached elements. Additionally, we propose software and hardware optimizations that designers can implement to create a platform suitable for sparse computation. We then investigate these optimizations using the gem5 simulator to achieve a significant speedup of up to 2.63× compared to a CPU where the optimizations are not applied., This research was supported by: the Spanish Ministry of Science and Innovation (MCIN) through contracts PID2019-107255GB-C21, PID2019-107255GBC21 (MCIN/AEI/10.13039/501100011033) and the European Processor Initiative (EPI) under grant agreement No. 101036168. M. Siracusa has been partially supported by an FI fellowship (2022FI B 00969). Finally, A. Armejach is a Serra Hunter Fellow, Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

A differential privacy protection-based federated deep learning framework to fog-embedded architectures

  • Gutiérrez Escobar, Norma
  • Otero Calviño, Beatriz|||0000-0002-9194-559X
  • Rodríguez Luna, Eva|||0000-0001-5904-7039
  • Utrera Iglesias, Gladys Miriam|||0000-0002-0637-1297
  • Mus León, Sergi
  • Canal Corretger, Ramon|||0000-0003-4542-204X
Nowadays, companies collect massive quantities of data to enhance their operations, often at the expense of sharing user sensible information. This data is widely used to train Deep Learning (DL) neural networks to model, classify, or recognize complex data. These activities enable companies to offer an array of services to users, such as precise advertising and optimal location services. This study explores potential solutions for preserving privacy while utilizing DL applications. To address the privacy issue, we develop a privacy-preserving framework specifically designed for fog computing environments. Unlike traditional cloud computing architectures, fog embedded architectures only share a small portion of user data with a nearby fog node, ensuring that the majority of sensitive data remains secure. Within these fog nodes, we incorporate two additional algorithms, namely Generalization and Threshold, to enhance the privacy-preserving capabilities of the framework. The first algorithm, Generalization, introduces a validation dataset within the fog nodes which not only increases the accuracy of the fog-embedded framework but also ensures that user data is preserved. The second algorithm, Threshold, is responsible for protecting user data samples and reducing the amount of information sent to the server. By combining these two algorithms, we are able to provide an additional layer of protection for user privacy while still maintaining the accuracy of the model. We conduct an evaluation to test its effectiveness using two separate datasets. In addition, we analyze them through a Feed Forward Neural Network (FFNN) and compare the results with a traditional centralized architecture to validate the effectiveness of the proposed framework. The results of our evaluation demonstrate that the proposed privacy-preserving framework, when combined with the Generalization and Threshold algorithms, can preserve up to 38.44% of user data. Additionally, we were able to extend the framework to multiple fog nodes without compromising the network’s accuracy, as we only observed a 0.1% decrease in accuracy when using the proposed architecture. This study emphasizes the importance of preserving user information while using DL applications and provides a solution that trains the desired network without violating user privacy, hence preserving their anonymity. Overall, the study highlights the potential of Federated Deep Learning to improve the accuracy and privacy of DL applications in fog computing environments., This work is partially supported by the Spanish Ministry of Science and Innovation, Spain under contracts PID2021-124463OB-IOO and PID2019-107255GB-C22; by the Generalitat de Catalunya, Spain under grant 2021SGR00326; and by the DRAC (IU16-011591), the HORIZON Vitamin-V, Spain(101093062), the HORIZON-AG PHOENI2X, Spain (101070586) and the HORIZON HORSE, Spain (101096342) projects., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

L2C2: Last-level compressed-contents non-volatile cache and a procedure to forecast performance and lifetime

  • Escuín Blasco, Carlos
  • Ibáñez Marín, Pablo
  • Navarro, Denis
  • Monreal Arnal, Teresa|||0000-0002-0458-2234
  • Llaberia Griñó, José M.|||0000-0002-3753-4108
  • Viñals Yúfera, Victor
Several emerging non-volatile (NV) memory technologies are rising as interesting alternatives to build the Last-Level Cache (LLC). Their advantages, compared to SRAM memory, are higher density and lower static power, but write operations wear out the bitcells to the point of eventually losing their storage capacity. In this context, this paper presents a novel LLC organization designed to extend the lifetime of the NV data array and a procedure to forecast in detail the capacity and performance of such an NV-LLC over its lifetime. From a methodological point of view, although different approaches are used in the literature to analyze the degradation of an NV-LLC, none of them allows to study in detail its temporal evolution. In this sense, this work proposes a forecasting procedure that combines detailed simulation and prediction, allowing an accurate analysis of the impact of different cache control policies and mechanisms (replacement, wear-leveling, compression, etc.) on the temporal evolution of the indices of interest, such as the effective capacity of the NV-LLC or the system IPC. We also introduce L2C2, a LLC design intended for implementation in NV memory technology that combines fault tolerance, compression, and internal write wear leveling for the first time. Compression is not used to store more blocks and increase the hit rate, but to reduce the write rate and increase the lifetime during which the cache supports near-peak performance. In addition, to support byte loss without performance drop, L2C2 inherently allows N redundant bytes to be added to each cache entry. Thus, L2C2+N, the endurance-scaled version of L2C2, allows balancing the cost of redundant capacity with the benefit of longer lifetime. For instance, as a use case, we have implemented the L2C2 cache with STT-RAM technology. It has affordable hardware overheads compared to that of a baseline NV-LLC without compression in terms of area, latency and energy consumption, and increases up to 6-37 times the time in which 50% of the effective capacity is degraded, depending on the variability in the manufacturing process. Compared to L2C2, L2C2+6 which adds 6 bytes of redundant capacity per entry, that means 9.1% of storage overhead, can increase up to 1.4-4.3 times the time in which the system gets its initial peak performance degraded., All authors acknowledge support in part from grants (1) PID2019-105660RB-C21, PID2019-107255GB-C22, PID2019-103939RB-I00, or PDC2021-120898-I00 from Agencia Estatal de Investigación (AEI, https://www.aei.gob.es/) and Ministerio de Ciencia e Innovación (MCIN, https://www.ciencia.gob.es), (2) gaZ: T5820R research group from Aragón Goverment (http://www.aragon.es), (3) EU through FEDER and NextGenerationEU/PRTR programs by the DGA-FSE, and (4) by ISCIII under Project PI21/00440., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

DDS: integrating data analytics transformations in task-based workflows [version 1, peer review: 1 approved, 2 approved with reservations]

  • Mammadli, Nihad
  • Ejarque Artigas, Jorge
  • Álvarez Cid-Fuentes, Javier
  • Badia Sala, Rosa Maria|||0000-0003-2941-5499
High-performance data analytics (HPDA) is a current trend in e-science research that aims to integrate traditional HPC with recent data analytic frameworks. Most of the work done in this field has focused on improving data analytic frameworks by implementing their engines on top of HPC technologies such as Message Passing Interface. However, there is a lack of integration from an application development perspective. HPC workflows have their own parallel programming models, while data analytic (DA) algorithms are mainly implemented using data transformations and executed with frameworks like Spark. Task-based programming models (TBPMs) are a very efficient approach for implementing HPC workflows. Data analytic transformations can also be decomposed as a set of tasks and implemented with a task-based programming model. In this paper, we present a methodology to develop HPDA applications on top of TBPMs that allow developers to combine HPC workflows and data analytic transformations seamlessly. A prototype of this approach has been implemented on top of the PyCOMPSs task- based programming model to validate two aspects: HPDA applications can be seamlessly developed and have better performance than Spark. We compare our results using different programs. Finally, we conclude with the idea of integrating DA into HPC applications and evaluation of our method against Spark., This research was financially supported by the European Union’s Horizon 2020 research and innovation programme under the grant agreement No 780622; and the Spanish Government (PID2019-107255GB), Generalitat de Catalunya (2014-SGR-1051)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Exploiting mixed-precision redundancy for soft-error detection in LU decomposition

  • Sahraneshinsamani, Nima
  • Catalán Pallarés, Sandra
  • Herrero Zaragoza, José Ramón|||0000-0002-4060-367X
Mixed-precision arithmetic, widely adopted for performance, creates an overlooked opportunity for fault tolerance. Our mixed-precision panel factorization (MPF) algorithm already computes each LU panel in both FP16 and FP64; we show that this inherent redundancy enables soft-error detection at negligible additional cost. We propose four complementary detectors that compare checksum sketches of the two factorizations: Relative (thresholded checksum discrepancy), Ratio (outlier analysis of per-row differences), Hybrid (a randomized-probe combination of both with separate L- and U-factor checks), and CrossCheck (a lightweight cross-check inspired by algorithm-based fault tolerance). Experimental evaluation shows that Relative and CrossCheck achieve high sensitivity with low false-positive rates, while Hybrid maintains near-zero false positives with moderate sensitivity and complementary coverage for challenging significand-bit faults. Bit-level analysis confirms reliable detection of exponent- and sign-bit errors, with sensitivity decreasing gracefully for low-order significand bits whose perturbations approach the FP16 rounding floor. A lightweight panel-fingerprint mechanism extends protection beyond the factorization loop, closing the temporal gap before the triangular solve with guaranteed detection of any single-element corruption and zero false positives. The approach requires only linear work and constant storage per panel, preserving the cubic scaling of standard LU decomposition., The authors would like to thank the Generalitat Valenciana for its support through project CIGE/2023/87. The researchers from UJI were additionally funded by project PID2023-146569NB-C22, supported by MICIU/AEI/10.13039/501100011033 and ERDF/EU. Sandra Catalán was supported by grant RYC2021-033973-I, funded by MCIN/AEI/10.13039/501100011033 and the European Union “NextGenerationEU”/PRTR. José R. Herrero was supported by project PID2019-107255GB of the Spanish Ministerio de Economía y Competitividad, and project 2021-SGR-01007 from the Generalitat de Catalunya., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Heterogeneous programming using OpenMP and CUDA/HIP for hybrid CPU-GPU scientific applications

  • González Tallada, Marc|||0000-0002-3780-1106
  • Morancho Llena, Enrique|||0000-0003-2403-8145
Hybrid computer systems combine compute units (CUs) of different nature like CPUs, GPUs and FPGAs. Simultaneously exploiting the computing power of these CUs requires a careful decomposition of the applications into balanced parallel tasks according to both the performance of each CU type and the communication costs among them. This paper describes the design and implementation of runtime support for OpenMP hybrid GPU-CPU applications, when mixed with GPU-oriented programming models (e.g. CUDA/HIP). The paper describes the case for a hybrid multi-level parallelization of the NPB-MZ benchmark suite. The implementation exploits both coarse-grain and fine-grain parallelism, mapped to compute units of different nature (GPUs and CPUs). The paper describes the implementation of runtime support to bridge OpenMP and HIP, introducing the abstractions of Computing Unit and Data Placement. We compare hybrid and non-hybrid executions under state-of-the-art schedulers for OpenMP: static and dynamic task schedulings. Then, we improve the set of schedulers with two additional variants: a memorizing-dynamic task scheduling and a profile-based static task scheduling. On a computing node composed of one AMD EPYC 7742 @ 2.250 GHz (64 cores and 2 threads/core, totalling 128 threads per node) and 2 × GPU AMD Radeon Instinct MI50 with 32 GB, hybrid executions present speedups from 1.10× up to 3.5× with respect to a non-hybrid GPU implementation, depending on the number of activated CUs., This work was supported by the Spanish Ministry of Science and Technology (PID2019-107255GB)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Acceleration strategies for large-scale sequential simulations using parallel neighbour search: Non-LVA and LVA scenarios

  • Peredo Andrade, Oscar Francisco|||0000-0002-1148-6312
  • Herrero Zaragoza, José Ramón|||0000-0002-4060-367X
This paper describes the application of acceleration techniques into existing implementations of Sequential Gaussian Simulation and Sequential Indicator Simulation. These implementations might incorporate Locally Varying Anisotropy (LVA) to capture non-linear features of the underlying physical phenomena. The imple- mentation focuses on a novel parallel neighbour search algorithm, which can be used on both non-LVA and LVA codes. Additionally, parallel shortest path executions and optimized linear algebra libraries are applied with focus on LVA codes. Execution time, speedup and accuracy results are presented. Non-LVA codes are benchmarked using two scenarios with approximately 50 million domain points each. Speedup results of 2× and 4× were obtained on SGS and SISIM respectively, where each scenario is compared against a baseline code published in Peredo et al. (2018). The aggregated contribution to speedup of both works results in 12× and 50× respectively. LVA codes are benchmarked using two scenarios with approximately 1.7 million domain points each. Speedup results of 56× and 1822× were obtained on SGS and SISIM respectively, where each scenario is compared against the original baseline sequential codes., The authors acknowledge the donated resources from project PID2019-107255GB of the Spanish Ministerio de Economía y Competitividad, and project 2017-SGR-1414 from the Generalitat de Catalunya, Spain., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Practicable live container migrations in high performance computing clouds: Diskless, iterative, and connection-persistent

  • Guitart Fernández, Jordi|||0000-0003-0751-3100
Checkpoint/Restore techniques had been thoroughly used by the High Performance Computing (HPC) community in the context of failure recovery. Given the current trend in HPC to use containerization to obtain fast, customized, portable, flexible, and reproducible deployments of their workloads, as well as efficient and reliable sharing and management of HPC Cloud infrastructures, there is a need to integrate Checkpoint/Restore with containerization in such a way that the freeze time of the application is minimal and live migrations are practicable. Whereas current Checkpoint/Restore tools (such as CRIU) support several options to accomplish this, most of them are rarely exploited in HPC Clouds and, consequently, their potential impact on the performance is barely known. Therefore, this paper explores the use of CRIU’s advanced features to implement diskless, iterative (pre-copy and post-copy) migrations of containers with external network namespaces and established TCP connections, so that memory-intensive and connection-persistent HPC applications can live-migrate. Our extensive experiments to characterize the performance impact of those features demonstrate that properly-configured live migrations incur low application downtime and memory/disk usage and are indeed feasible in containerized HPC Clouds., This research was partially supported by the Spanish Government under contract PID2019-107255GB-C22, by the Generalitat de Catalunya, Spain under contract 2021-SGR-00478, and by the EU-HORIZON programme under grant agreement 101092646., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Multi-GPU systems and Unified Virtual Memory for scientific applications: The case of the NAS multi-zone parallel benchmarks

  • González Tallada, Marc|||0000-0002-3780-1106
  • Morancho Llena, Enrique|||0000-0003-2403-8145
GPU-based computing systems have become a widely accepted solution for the high-performance-computing (HPC) domain. GPUs have shown highly competitive performance-per-watt ratios and can exploit an astonishing level of parallelism. However, exploiting the peak performance of such devices is a challenge, mainly due to the combination of two essential aspects of multi-GPU execution: memory allocation and work distribution. Memory allocation determines the data mapping to GPUs, and therefore conditions all work distribution schemes and communication phases in the application. Unified Virtual Memory simplifies the codification of memory allocations, but its effects on performance depend on how data is used by the devices and how the devices' driver is going to orchestrate the data transfers across the system. In this paper we present a multi-GPU and Unified Virtual Memory (UM) implementation of the NAS Multi-Zone Parallel Benchmarks which alternate communication and computation phases offering opportunities to overlap these phases. We analyse the programmability and performance effects of the introduction of the UM support. Our experience shows that the programming efforts for introducing UM are similar to those of having a memory allocation per GPU. On an evaluation environment composed of 2 x IBM Power9 8335-GTH and 4 x GPU NVIDIA V100 (Volta), our UM-based parallelization outperforms the manual memory allocation versions by 1.10x to 1.85x. However, these improvements are highly sensitive to the information forwarded to the devices' driver describing the most convenient location for specific memory regions. We analyse these improvements in terms of the relationship between the computational and communication phases of the applications., This work was supported by the Spanish Ministry of Science and Technology (PID2019-107255GB)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Exploring the potential of BCI in education: an experiment in musical training

  • Folgieri, Raffaella
  • Lucchiari, Claudio
  • Gricar, Sergej
  • Baldigara, Tea
  • Gil, Marisa|||0000-0001-9755-3311
Brain–computer interfaces (BCIs) have gained significant attention in recent years for various applications, including education and skill development: studies have shown that BCIs can boost memory, concentration, and even creativity and can improve learning and memory retention in healthy people. In our current study, we investigated the effectiveness of real-time feedback provided by a BCI system for improving performance on a specific task. A total of 20 participants completed a pre-training assessment, followed by a training period with the BCI system and a post-training assessment. The BCI system provided real-time feedback based on the participants’ level of accuracy, with positive feedback given for scores above 70%. Results showed a significant improvement in accuracy scores from pre- to post-training, with an average improvement of 15%. Participants also reported high levels of satisfaction with the feedback provided by the BCI system. These findings suggest that real-time feedback provided by a BCI system can be an effective tool for skill development and education, particularly when tailored to the specific needs of individual learners. Further research is needed to explore the potential of BCIs for a wide range of educational applications., This research was funded by the Department of Philosophy “Piero Martinetti” of the University of Milan under the Project “Departments of Excellence 2023–2027” awarded by the Ministry of University and Research (MUR). Marisa Gil’s work has been supported by the Spanish Ministry of Education (PID2019-107255GB-C22) and partially supported by the Generalitat de Catalunya (contract 2021-SGR-01007). Certain sections of this article were written using tools funded by the Slovenian Research and Innovation Agency under grant number CRP2023 V5-2331. The same institutions funded the APC., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Automated generation of high-performance computational fluid dynamics codes

  • Macià Sorrosal, Sandra
  • Martínez Ferrer, Pedro J.
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
  • Beltran Querol, Vicenç
Domain-Specific Languages (DSLs) improve programmers productivity by decoupling problem descriptions from algorithmic implementations. However, DSLs for High-Performance Computing (HPC) have two additional critical requirements: performance and scalability. This paper presents the automated process of generating, from abstract mathematical specifications of Computational Fluid Dynamics (CFD) problems, optimised parallel codes that perform and scale as manually optimised ones. We consciously combine within Saiph, a DSL for solving CFD problems, low-level optimisations and parallelisation strategies, enabling high-performance single-core executions which effectively scale to multi-core and distributed environments. Our results demonstrate how high-level DSLs can offer competitive performance by transparently leveraging state-of-the-art HPC techniques., This research has received funding from the European Union’s Horizon 2020/EuroHPC research and innovation programme under grant agreement N. 955606 (DEEP-SEA), and is supported by the Spanish State Research Agency - Ministry of Science and Innovation (contract PID2019-107255 GB), and by the Generalitat de Catalunya (2017-SGR1414). This work is also supported by the Ministry of Economy of Spain through Severo Ochoa Center of Excellence Program (SEV-2015-0493)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Lester: rotoscope animation through video object segmentation and tracking

  • Tous Liesa, Rubén|||0000-0002-1409-5843
This article introduces Lester, a novel method to automatically synthesize retro-style 2D animations from videos. The method approaches the challenge mainly as an object segmentation and tracking problem. Video frames are processed with the Segment Anything Model (SAM) and the resulting masks are tracked through subsequent frames with DeAOT, a method of hierarchical propagation for semi-supervised video object segmentation. The geometry of the masks’ contours is simplified with the Douglas–Peucker algorithm. Finally, facial traits, pixelation and a basic rim light effect can be optionally added. The results show that the method exhibits an excellent temporal consistency and can correctly process videos with different poses and appearances, dynamic shots, partial shots and diverse backgrounds. The proposed method provides a more simple and deterministic approach than diffusion models based video-to-video translation pipelines, which suffer from temporal consistency problems and do not cope well with pixelated and schematic outputs. The method is also more feasible than techniques based on 3D human pose estimation, which require custom handcrafted 3D models and are very limited with respect to the type of scenes they can process., This research is partially funded by the Spanish Ministry of Science and Innovation under contract PID2019-107255GB, and by the SGR programme 2021-SGR-00478 of the Catalan Government., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Role-shifting threads: Increasing OpenMP malleability to address load imbalance at MPI and OpenMP

  • Criado Ledesma, Joel
  • López Herrero, Víctor
  • Vinyals Ylla Català, Joan
  • Ramirez Miranda, Guillem
  • Teruel García, Xavier
  • Garcia Gasulla, Marta|||0000-0003-3682-9905
This paper presents the evolution of the free agent threads for OpenMP to the new role-shifting threads model and their integration with the Dynamic Load Balancing (DLB) library. We demonstrate how free agent threads can improve resource utilization in OpenMP applications with load imbalance in their nested parallel regions. We also demonstrate how DLB efficiently manages the malleability exposed by the role-shifting threads to address load imbalance issues. We use three real-world scientific applications, one of them to demonstrate that free agents alone can improve the OpenMP model without external tools, and two other MPI+OpenMP applications, one of them with a coupling case, to illustrate the potential of the free agent threads’ malleability with an external resource manager to increase the efficiency of the system. In addition, we demonstrate that the new implementation is more usable than the former one, letting the runtime system automatically make decisions that were made by the programmer previously. All software is released open-source., This work has received funding from the DEEP Projects, at the European Commission’s FP7, H2020, and EuroHPC Programmes, under Grant Agreements 287530, 610476, 754304, and 955606. The PCI2021-121958 financed by the Spanish State Research Agency - Ministry of Science and Innovation. And it also has the support of the Spanish Ministry of Science and Innovation (Computacion de Altas Prestaciones VIII: PID2019-107255GB)., Peer Reviewed
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Role-shifting threads: Increasing OpenMP malleability to address load imbalance at MPI and OpenMP

  • Criado Ledesma, Joel
  • López Herrero, Víctor
  • Vinyals Ylla Català, Joan
  • Ramirez Miranda, Guillem
  • Teruel García, Xavier
  • Garcia Gasulla, Marta
This paper presents the evolution of the free agent threads for OpenMP to the new role-shifting threads model and their integration with the Dynamic Load Balancing (DLB) library. We demonstrate how free agent threads can improve resource utilization in OpenMP applications with load imbalance in their nested parallel regions. We also demonstrate how DLB efficiently manages the malleability exposed by the role-shifting threads to address load imbalance issues. We use three real-world scientific applications, one of them to demonstrate that free agents alone can improve the OpenMP model without external tools, and two other MPI+OpenMP applications, one of them with a coupling case, to illustrate the potential of the free agent threads’ malleability with an external resource manager to increase the efficiency of the system. In addition, we demonstrate that the new implementation is more usable than the former one, letting the runtime system automatically make decisions that were made by the programmer previously. All software is released open-source., This work has received funding from the DEEP Projects, at the European Commission’s FP7, H2020, and EuroHPC Programmes, under Grant Agreements 287530, 610476, 754304, and 955606. The PCI2021-121958 financed by the Spanish State Research Agency - Ministry of Science and Innovation. And it also has the support of the Spanish Ministry of Science and Innovation (Computacion de Altas Prestaciones VIII: PID2019-107255GB)., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Boosting HPC data analysis performance with the ParSoDA-Py library

  • Belcastro, Loris
  • Giampà, Salvatore
  • Marozzo, Fabrizio
  • Talia, Domenico
  • Trunfio, Paolo
  • Badia Sala, Rosa Maria
  • Ejarque, Jorge
  • Mammadli, Nihad
Developing and executing large-scale data analysis applications in parallel and distributed environments can be a complex and time-consuming task. Developers often find themselves diverted from their application logic to handle technical details about the underlying runtime and related issues. To simplify this process, ParSoDA, a Java library, has been proposed to facilitate the development of parallel data mining applications executed on HPC systems. It simplifies the process by providing built-in scalability mechanisms relying on the Hadoop and Spark frameworks. This paper presents ParSoDA-Py, the Python version of the ParSoDA library, which allows for further support of commonly used runtimes and libraries for big data analysis. After a complete library redesign, ParSoDA can be now easily integrated with other Python-based distributed runtimes for HPC systems, such as COMPSs and Apache Spark, and with the large ecosystem of Python-based data processing libraries. The paper discusses the adaptation process, which takes into consideration the new technical requirements, and evaluates both usability and scalability through some case study applications., This work has been partially funded by the European Commission’s Horizon 2020 Framework program and the European High-Performance Computing Joint Undertaking (JU) under Grant agreement No 955558 and by MCIN/AEI/10.13039/501100011033 and the European Union NextGenerationEU/PRTR (PCI2021-121957), project eFlows4HPC. It has also been supported by the Spanish Government (PID2019-107255GB) and by the Departament de Recerca i Universitats de la Generalitat de Catalunya to the Research Group MPiEDist (2021 SGR 00412) We also acknowledge financial support from “National Centre for HPC, Big Data and Quantum Computing," CN00000013 - CUP H23C22000360005, and from “FAIR - Future Artificial Intelligence Research" Project - CUP H23C22000860006., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Programming parallel dense matrix factorizations and inversion for new-generation NUMA architectures

  • Catalán Pallarés, Sandra
  • Igual Peña, Francisco D.
  • Herrero Zaragoza, José Ramón
  • Rodríguez Sánchez, Rafael
  • Quintana Ortí, Enrique Salvador
We propose a methodology to address the programmability issues derived from the emergence of new-generation shared-memory NUMA architectures. For this purpose, we employ dense matrix factorizations and matrix inversion (DMFI) as a use case, and we target two modern architectures (AMD Rome and Huawei Kunpeng 920) that exhibit configurable NUMA topologies. Our methodology pursues performance portability across different NUMA configurations by proposing multi-domain implementations for DMFI plus a hybrid task- and loop-level parallelization that configures multi-threaded executions to fix core-to-data binding, exploiting locality at the expense of minor code modifications. In addition, we introduce a generalization of the multi-domain implementations for DMFI that offers support for virtually any NUMA topology in present and future architectures. Our experimentation on the two target architectures for three representative dense linear algebra operations validates the proposal, reveals insights on the necessity of adapting both the codes and their execution to improve data access locality, and reports performance across architectures and inter- and intra-socket NUMA configurations competitive with state-of-the-art message-passing implementations, maintaining the ease of development usually associated with shared-memory programming., This research was sponsored by project PID2019-107255GB of Ministerio de Ciencia, Innovación y Universidades; project S2018/TCS-4423 of Comunidad de Madrid; project 2017-SGR-1414 of the Generalitat de Catalunya and the Madrid Government under the Multiannual Agreement with UCM in the line Program to Stimulate Research for Young Doctors in the context of the V PRICIT, project PR65/19-22445. This project has also received funding from the European High-Performance Computing Joint Undertaking (JU) under grant agreement No 955558. The JU receives support from the European Union’s Horizon 2020 research and innovation programme, and Spain, Germany, France, Italy, Poland, Switzerland, Norway. The work is also supported by grants PID2020-113656RB-C22 and PID2021-126576NB-I00 of MCIN/AEI/10.13039/501100011033 and by ERDF A way of making Europe., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Automated generation of high-performance computational fluid dynamics codes

  • Macià Sorrosal, Sandra
  • Martínez Ferrer, Pedro J.
  • Ayguadé Parra, Eduard
  • Beltran Querol, Vicenç
Domain-Specific Languages (DSLs) improve programmers productivity by decoupling problem descriptions from algorithmic implementations. However, DSLs for High-Performance Computing (HPC) have two additional critical requirements: performance and scalability. This paper presents the automated process of generating, from abstract mathematical specifications of Computational Fluid Dynamics (CFD) problems, optimised parallel codes that perform and scale as manually optimised ones. We consciously combine within Saiph, a DSL for solving CFD problems, low-level optimisations and parallelisation strategies, enabling high-performance single-core executions which effectively scale to multi-core and distributed environments. Our results demonstrate how high-level DSLs can offer competitive performance by transparently leveraging state-of-the-art HPC techniques., This research has received funding from the European Union’s Horizon 2020/EuroHPC research and innovation programme under grant agreement N. 955606 (DEEP-SEA), and is supported by the Spanish State Research Agency - Ministry of Science and Innovation (contract PID2019-107255 GB), and by the Generalitat de Catalunya (2017-SGR1414). This work is also supported by the Ministry of Economy of Spain through Severo Ochoa Center of Excellence Program (SEV-2015-0493)., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Low-cost training of image-to-image diffusion models with incremental learning and task/domain adaptation

  • Antona Pizà, Héctor
  • Otero Calviño, Beatriz
  • Tous Liesa, Rubén
Diffusion models specialized in image-to-image translation tasks, like inpainting and colorization, have outperformed the state of the art, yet their computational requirements are exceptionally demanding. This study analyzes different strategies to train image-to-image diffusion models in a low-resource setting. The studied strategies include incremental learning and task/domain transfer learning. First, a base model for human face inpainting is trained from scratch with an incremental learning strategy. The resulting model achieves an FID score almost equivalent to that of its batch learning equivalent while significantly reducing the training time. Second, the base model is fine-tuned to perform a different task, image colorization, and, in a different domain, landscape images. The resulting colorization models showcase exceptional performances with a minimal number of training epochs. We examine the impact of different configurations and provide insights into the ability of image-to-image diffusion models for transfer learning across tasks and domains., This work is partially supported by the Spanish Ministry of Science and Innovation under contracts PID2019-107255GB and PID2021-124463OB-IOO, by the Generalitat de Catalunya under grants 2017-SGR-962, 2021-SGR-00326, 2021-SGR-00478 and DRAC (IU16-011591). The research leading to these results has received funding from the European Union’s Horizon 2020 research and innovation program under the HORIZON-AG PHOENI2X (101070586) and HORIZON-EU VITAMIN-V (101093062) projects., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

SpChar: characterizing the sparse puzzle via decision trees

  • Sgherzi, Francesco
  • Siracusa, Marco
  • Fernández Vega, Ivan
  • Armejach Sanosa, Adrià
  • Moretó Planas, Miquel
Sparse matrix computation is crucial in various modern applications, including large-scale graph analytics, deep learning, and recommender systems. The performance of sparse kernels varies greatly depending on the structure of the input matrix, making it difficult to gain a comprehensive understanding of sparse computation and its relationship to inputs, algorithms, and target machine architecture. Despite extensive research on certain sparse kernels, such as Sparse Matrix-Vector Multiplication (SpMV), the overall family of sparse algorithms has yet to be investigated as a whole. This paper introduces SpChar, a workload characterization methodology for general sparse computation. SpChar employs tree-based models to identify the most relevant hardware and input characteristics, starting from hardware and input-related metrics gathered from Performance Monitoring Counters (PMCs) and matrices. Our analysis enables the creation of a characterization loop that facilitates the optimization of sparse computation by mapping the impact of architectural features to inputs and algorithmic choices. We apply SpChar to more than 600 matrices from the SuiteSparse Matrix collection and three state-of-the-art Arm Central Processing Units (CPUs) to determine the critical hardware and software characteristics that affect sparse computation. In our analysis, we determine that the biggest limiting factors for high-performance sparse computation are (1) the latency of the memory system, (2) the pipeline flush overhead resulting from branch misprediction, and (3) the poor reuse of cached elements. Additionally, we propose software and hardware optimizations that designers can implement to create a platform suitable for sparse computation. We then investigate these optimizations using the gem5 simulator to achieve a significant speedup of up to 2.63× compared to a CPU where the optimizations are not applied., This research was supported by: the Spanish Ministry of Science and Innovation (MCIN) through contracts PID2019-107255GB-C21, PID2019-107255GBC21 (MCIN/AEI/10.13039/501100011033) and the European Processor Initiative (EPI) under grant agreement No. 101036168. M. Siracusa has been partially supported by an FI fellowship (2022FI B 00969). Finally, A. Armejach is a Serra Hunter Fellow, Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Automatic distributed deep learning using resource-constrained edge devices

  • Gutiérrez Torre, Alberto
  • Bahadori, Kiyana
  • Baig, Shuja-ur-Rehman
  • Iqbal, Waheed
  • Vardanega, Tullio
  • Berral García, Josep Lluís
  • Carrera Pérez, David
Processing data generated at high volume and speed from the Internet of Things, smart cities, domotic, intelligent surveillance, and e-healthcare systems require efficient data processing and analytics services at the Edge to reduce the latency and response time of the applications. The Fog Computing Edge infrastructure consists of devices with limited computing, memory, and bandwidth resources, which challenge the construction of predictive analytics solutions that require resource-intensive tasks for training machine learning models. In this work, we focus on the development of predictive analytics for urban traffic. Our solution is based on deep learning techniques localized in the Edge, where computing devices have very limited computational resources. We present an innovative method for efficiently training of Gated Recurrent-Units (GRUs) across available resource-constrained CPU and GPU Edge devices. Our solution employs distributed GRU model learning and dynamically stops the training process to utilize the low-power and resource-constrained Edge devices while ensuring good estimation accuracy effectively. The proposed solution was extensively evaluated using low-powered ARM-based devices, including Raspberry Pi v3 and the low-powered GPU-enabled device NVIDIA Jetson Nano, and also compared them with Single-CPU Intel Xeon machines. For the evaluation experiments, we used real-world Floating Car Data. The experiments show that the proposed solution delivers excellent prediction accuracy and computational performance on the Edge when compared with the baseline methods., This work has been partially supported by the Spanish Government (contract PID2019-107255GB), the Generalitat de Catalunya (contract 2014-SGR-1051), the University of Padova, and Severo Ochoa CoE (SEV-2015-0493-16-5). The statements made herein are solely the responsibility of the authors., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

A closer look at referring expressions for video object segmentation

  • Bellver Bueno, Míriam
  • Ventura Royo, Carles
  • Silberer, Carina
  • Kazakos, Ioannis
  • Torres Viñals, Jordi
  • Giró Nieto, Xavier
The task of Language-guided Video Object Segmentation (LVOS) aims at generating binary masks for an object referred by a linguistic expression. When this expression unambiguously describes an object in the scene, it is named referring expression (RE). Our work argues that existing benchmarks used for LVOS are mainly composed of trivial cases, in which referents can be identified with simple phrases. Our analysis relies on a new categorization of the referring expressions in the DAVIS-2017 and Actor-Action datasets into trivial and non-trivial REs, where the non-trivial REs are further annotated with seven RE semantic categories. We leverage these data to analyze the performance of RefVOS, a novel neural network that obtains competitive results for the task of language-guided image segmentation and state of the art results for LVOS. Our study indicates that the major challenges for the task are related to understanding motion and static actions., Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature. This work was partially supported by the projects PID2019-107255GB-C22 and PID2020-117142GB-I00 funded by MCIN/ AEI /10.13039/501100011033 Spanish Ministry of Science, and the grant 2017-SGR-1414 of the Government of Catalonia. This work was also partially supported by the project RTI2018-095232-B-C22 funded by the Spanish Ministry of Science, Innovation and Universities., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Human pose completion in partial body camera shots

  • Tous Liesa, Rubén
  • Nin Guerrero, Jordi
  • Igual Muñoz, Laura
Many actual images contain partial body camera shots, in which a significant part of the body is not visible. This issue is especially prevalent in film images, where less than 10% are full-body shots. Most 2D human pose estimation methods return incomplete poses when applied to partial body images. This lack of completeness becomes a problem in some situations, for example, when the 2D pose is converted to a 3D pose by a two-stage 3D human pose estimation method since most of these methods require a complete pose to work. This article proposes a new technique, called CompletePose, consisting of completing the missing keypoints when 2D human pose estimation methods are applied to images with partial body camera shots. A Conditional Generative Adversarial Network is used to obtain a complete and plausible pose, realistic enough to predict a 3D pose with a two-stage 3D human pose estimation method. A complete empirical validation has been carried out with the Human3.6 M dataset and a new dataset, called CHARADE, specially built and made public for reproducibility and benchmarking for this research. Quantitative evaluation employing the Fréchet distance shows that the approach manages to approximate the actual data distribution. The qualitative evaluation shows that the completed poses enable obtaining plausible 3D poses from images previously intractable., This work is partially supported by the Spanish Ministry of Science and Innovation under contracts PID2019-107255GB, PLEC2021-007850 and RTI2018-095232-B-C21, and by the SGR programme of the Catalan Government under contracts 2021-SGR00478, 2017-SGR-1414 and 2017-SGR-1742., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Pictonaut: movie cartoonization using 3D human pose estimation and GANs

  • Tous Liesa, Rubén
This article describes Pictonaut, a novel method to automatically synthetise animated shots from motion picture footage. Its results are editable (backgrounds, characters, lighting, etc.) with conventional 3D software, and they have the finish of professional 2D animation. Rather than addressing the challenge solely as an image translation problem, a hybrid approach combining multi-person 3D human pose estimation and GANs is taken. Sub-sampled video frames are processed with OpenPose and SMPLify-X to obtain the 3D parameters of the pose (body, hands and face expression) of all depicted characters. The captured parameters are retargeted into manually selected 3D models, cel shaded to mimic the style of a 2D cartoon. The results of sub-sampled frames are interpolated to generate a complete and smooth motion for all the characters. The background is cartoonized with a GAN. Qualitative evaluation shows that the approach is feasible, and a small dataset of synthetised shots obtained from real movie scenes is provided., This work is partially supported by the Spanish Ministry of Science and Innovation under contract PID2019-107255GB, and by the SGR programme 2017-SGR-1414 of the Catalan Government., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Evaluation of work distribution schedulers for heterogeneous architectures and scientific applications

  • González Tallada, Marc
  • Morancho Llena, Enrique
This article explores and evaluates variants of state-of-the-art work distribution schemes adapted for scientific applications running on hybrid systems. A hybrid implementation (multi-GPU and multi-CPU) of the NASA Parallel Benchmarks - MutiZone (NPB-MZ) benchmarks is described to study the different elements that condition the execution of this suite of applications when parallelism is spread over a set of computing units (CUs) of different computational power (e.g., GPUs and CPUs). This article studies the influence of the work distribution schemes on the data placement across the devices and the host, which in turn determine the communications between the CUs, and evaluates how the schedulers are affected by the relationship between data placement and communications. We show that only the schedulers aware of the different computational power of the CUs and minimize communications are able to achieve an appropriate work balance and high performance levels. Only then does the combination of GPUs and CPUs result in an effective parallel implementation that boosts the performance of a non-hybrid multi-GPU implementation. The article describes and evaluates the schedulers static-pcf, Guided, and Clustered Guided to solve the previously mentioned limitations that appear in hybrid systems. We compare them against state-of-the-art static and memorizing dynamic schedulers. Finally, on a system with an AMD EPYC 7742 at 2.250GHz (64 cores, 2 threads per core, 128 threads) and two AMD Radeon Instinct MI50 GPUs with 32GB, we have observed that hybrid executions speed up from 1.1× up to 3.5× with respect to a non-hybrid GPU implementation., This work was supported by the Spanish Ministry of Science and Technology under the grant no. PID2019-107255GB., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Multi-GPU systems and Unified Virtual Memory for scientific applications: The case of the NAS multi-zone parallel benchmarks

  • González Tallada, Marc
  • Morancho Llena, Enrique
GPU-based computing systems have become a widely accepted solution for the high-performance-computing (HPC) domain. GPUs have shown highly competitive performance-per-watt ratios and can exploit an astonishing level of parallelism. However, exploiting the peak performance of such devices is a challenge, mainly due to the combination of two essential aspects of multi-GPU execution: memory allocation and work distribution. Memory allocation determines the data mapping to GPUs, and therefore conditions all work distribution schemes and communication phases in the application. Unified Virtual Memory simplifies the codification of memory allocations, but its effects on performance depend on how data is used by the devices and how the devices' driver is going to orchestrate the data transfers across the system. In this paper we present a multi-GPU and Unified Virtual Memory (UM) implementation of the NAS Multi-Zone Parallel Benchmarks which alternate communication and computation phases offering opportunities to overlap these phases. We analyse the programmability and performance effects of the introduction of the UM support. Our experience shows that the programming efforts for introducing UM are similar to those of having a memory allocation per GPU. On an evaluation environment composed of 2 x IBM Power9 8335-GTH and 4 x GPU NVIDIA V100 (Volta), our UM-based parallelization outperforms the manual memory allocation versions by 1.10x to 1.85x. However, these improvements are highly sensitive to the information forwarded to the devices' driver describing the most convenient location for specific memory regions. We analyse these improvements in terms of the relationship between the computational and communication phases of the applications., This work was supported by the Spanish Ministry of Science and Technology (PID2019-107255GB)., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

L2C2: Last-level compressed-contents non-volatile cache and a procedure to forecast performance and lifetime

  • Escuín Blasco, Carlos
  • Ibáñez Marín, Pablo
  • Navarro, Denis
  • Monreal Arnal, Teresa
  • Llaberia Griñó, José M.
  • Viñals Yúfera, Victor
Several emerging non-volatile (NV) memory technologies are rising as interesting alternatives to build the Last-Level Cache (LLC). Their advantages, compared to SRAM memory, are higher density and lower static power, but write operations wear out the bitcells to the point of eventually losing their storage capacity. In this context, this paper presents a novel LLC organization designed to extend the lifetime of the NV data array and a procedure to forecast in detail the capacity and performance of such an NV-LLC over its lifetime. From a methodological point of view, although different approaches are used in the literature to analyze the degradation of an NV-LLC, none of them allows to study in detail its temporal evolution. In this sense, this work proposes a forecasting procedure that combines detailed simulation and prediction, allowing an accurate analysis of the impact of different cache control policies and mechanisms (replacement, wear-leveling, compression, etc.) on the temporal evolution of the indices of interest, such as the effective capacity of the NV-LLC or the system IPC. We also introduce L2C2, a LLC design intended for implementation in NV memory technology that combines fault tolerance, compression, and internal write wear leveling for the first time. Compression is not used to store more blocks and increase the hit rate, but to reduce the write rate and increase the lifetime during which the cache supports near-peak performance. In addition, to support byte loss without performance drop, L2C2 inherently allows N redundant bytes to be added to each cache entry. Thus, L2C2+N, the endurance-scaled version of L2C2, allows balancing the cost of redundant capacity with the benefit of longer lifetime. For instance, as a use case, we have implemented the L2C2 cache with STT-RAM technology. It has affordable hardware overheads compared to that of a baseline NV-LLC without compression in terms of area, latency and energy consumption, and increases up to 6-37 times the time in which 50% of the effective capacity is degraded, depending on the variability in the manufacturing process. Compared to L2C2, L2C2+6 which adds 6 bytes of redundant capacity per entry, that means 9.1% of storage overhead, can increase up to 1.4-4.3 times the time in which the system gets its initial peak performance degraded., All authors acknowledge support in part from grants (1) PID2019-105660RB-C21, PID2019-107255GB-C22, PID2019-103939RB-I00, or PDC2021-120898-I00 from Agencia Estatal de Investigación (AEI, https://www.aei.gob.es/) and Ministerio de Ciencia e Innovación (MCIN, https://www.ciencia.gob.es), (2) gaZ: T5820R research group from Aragón Goverment (http://www.aragon.es), (3) EU through FEDER and NextGenerationEU/PRTR programs by the DGA-FSE, and (4) by ISCIII under Project PI21/00440., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Lester: rotoscope animation through video object segmentation and tracking

  • Tous Liesa, Rubén
This article introduces Lester, a novel method to automatically synthesize retro-style 2D animations from videos. The method approaches the challenge mainly as an object segmentation and tracking problem. Video frames are processed with the Segment Anything Model (SAM) and the resulting masks are tracked through subsequent frames with DeAOT, a method of hierarchical propagation for semi-supervised video object segmentation. The geometry of the masks’ contours is simplified with the Douglas–Peucker algorithm. Finally, facial traits, pixelation and a basic rim light effect can be optionally added. The results show that the method exhibits an excellent temporal consistency and can correctly process videos with different poses and appearances, dynamic shots, partial shots and diverse backgrounds. The proposed method provides a more simple and deterministic approach than diffusion models based video-to-video translation pipelines, which suffer from temporal consistency problems and do not cope well with pixelated and schematic outputs. The method is also more feasible than techniques based on 3D human pose estimation, which require custom handcrafted 3D models and are very limited with respect to the type of scenes they can process., This research is partially funded by the Spanish Ministry of Science and Innovation under contract PID2019-107255GB, and by the SGR programme 2021-SGR-00478 of the Catalan Government., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Exploiting mixed-precision redundancy for soft-error detection in LU decomposition

  • Sahraneshinsamani, Nima
  • Catalán Pallarés, Sandra
  • Herrero Zaragoza, José Ramón
Mixed-precision arithmetic, widely adopted for performance, creates an overlooked opportunity for fault tolerance. Our mixed-precision panel factorization (MPF) algorithm already computes each LU panel in both FP16 and FP64; we show that this inherent redundancy enables soft-error detection at negligible additional cost. We propose four complementary detectors that compare checksum sketches of the two factorizations: Relative (thresholded checksum discrepancy), Ratio (outlier analysis of per-row differences), Hybrid (a randomized-probe combination of both with separate L- and U-factor checks), and CrossCheck (a lightweight cross-check inspired by algorithm-based fault tolerance). Experimental evaluation shows that Relative and CrossCheck achieve high sensitivity with low false-positive rates, while Hybrid maintains near-zero false positives with moderate sensitivity and complementary coverage for challenging significand-bit faults. Bit-level analysis confirms reliable detection of exponent- and sign-bit errors, with sensitivity decreasing gracefully for low-order significand bits whose perturbations approach the FP16 rounding floor. A lightweight panel-fingerprint mechanism extends protection beyond the factorization loop, closing the temporal gap before the triangular solve with guaranteed detection of any single-element corruption and zero false positives. The approach requires only linear work and constant storage per panel, preserving the cubic scaling of standard LU decomposition., The authors would like to thank the Generalitat Valenciana for its support through project CIGE/2023/87. The researchers from UJI were additionally funded by project PID2023-146569NB-C22, supported by MICIU/AEI/10.13039/501100011033 and ERDF/EU. Sandra Catalán was supported by grant RYC2021-033973-I, funded by MCIN/AEI/10.13039/501100011033 and the European Union “NextGenerationEU”/PRTR. José R. Herrero was supported by project PID2019-107255GB of the Spanish Ministerio de Economía y Competitividad, and project 2021-SGR-01007 from the Generalitat de Catalunya., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Adaptive power shifting for power-constrained heterogeneous systems

  • Ortega Carrasco, Cristobal
  • Alvarez Martí, Lluc
  • Buyuktosunoglu, Alper
  • Bertran Monfort, Ramon
  • Rosedahl, Todd Jon
  • Bose, Pradip
  • Moretó Planas, Miquel
The number and heterogeneity of compute devices, even within a single compute node, has been steadily on the rise. Since all systems must operate under a power cap, the number of discrete devices that can run simultaneously at their highest frequency is limited by the globally-imposed power cap. Current systems incorporate a centralized power management unit that statically controls the distribution of power among the devices within the node. However, such static distribution policies are unaware of the dynamic utilization profile across the devices, which leads to unfair power allocations that end up degrading system throughput performance. The problem is particularly acute in the presence of heterogeneity since type-specific performance-boost capabilities cannot be leveraged via utilization-agnostic static power allocations. This paper proposes Adaptive Power Shifting for multi-accelerator heterogeneous systems (APS), a technique that leverages system utilization information to dynamically allocate and re-distribute power budgets across multiple discrete devices. Democratizing the power allocation based on dynamic needs results in dramatic speedup over a need-agnostic static allocation. We use APS in a real OpenPOWER compute node with 2 CPUs and 4 GPUs to demonstrate the value of on-demand, equitable power allocations. Overall, the proposed solution increases performance with respect to two state-of-the-art techniques by up to 14.9% and 13.8%., This work has been partially supported by the European Union’s Horizon 2020 research and innovation program under the Mont-Blanc 2020 project (grant agreement 779877), by the Spanish Ministry of Science and Innovation (contract PID2019-107255GB-C22), by Generalitat de Catalunya (contracts 2014-SGR-1051 and 2014-SGR-1272) and by the IBM/BSC Deep Learning Center initiative. Ll. Alvarez has been supported in part by the Spanish Ministry of Economy, Industry and Competitiveness under the Juan de la Cierva Formacion fellowship No. FJCI-2016- 30984. M. Moreto has been supported in part by the Spanish Ministry of Economy, Industry and Competitiveness under Ramon y Cajal fellowship No. RYC-2016-21104., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

DDS: integrating data analytics transformations in task-based workflows [version 1; peer review: 1 approved, 2 approved with reservations]

  • Mammadli, Nihad
  • Ejarque Artigas, Jorge
  • Álvarez Cid-Fuentes, Javier
  • Badia Sala, Rosa Maria
High-performance data analytics (HPDA) is a current trend in e-science research that aims to integrate traditional HPC with recent data analytic frameworks. Most of the work done in this field has focused on improving data analytic frameworks by implementing their engines on top of HPC technologies such as Message Passing Interface. However, there is a lack of integration from an application development perspective. HPC workflows have their own parallel programming models, while data analytic (DA) algorithms are mainly implemented using data transformations and executed with frameworks like Spark. Task-based programming models (TBPMs) are a very efficient approach for implementing HPC workflows. Data analytic transformations can also be decomposed as a set of tasks and implemented with a task-based programming model. In this paper, we present a methodology to develop HPDA applications on top of TBPMs that allow developers to combine HPC workflows and data analytic transformations seamlessly. A prototype of this approach has been implemented on top of the PyCOMPSs task- based programming model to validate two aspects: HPDA applications can be seamlessly developed and have better performance than Spark. We compare our results using different programs. Finally, we conclude with the idea of integrating DA into HPC applications and evaluation of our method against Spark., This research was financially supported by the European Union’s Horizon 2020 research and innovation programme under the grant agreement No 780622; and the Spanish Government (PID2019-107255GB), Generalitat de Catalunya (2014-SGR-1051)., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Performance characterization of containerization for HPC workloads on InfiniBand clusters: an empirical study

  • Liu, Peini
  • Guitart Fernández, Jordi
Containerization technology offers an appealing alternative for encapsulating and operating applications (and all their dependencies) without being constrained by the performance penalties of using Virtual Machines and, as a result, has got the interest of the High-Performance Computing (HPC) community to obtain fast, customized, portable, flexible, and reproducible deployments of their workloads. Previous work on this area has demonstrated that containerized HPC applications can exploit InfiniBand networks, but has ignored the potential of multi-container deployments which partition the processes that belong to each application into multiple containers in each host. Partitioning HPC applications has demonstrated to be useful when using virtual machines by constraining them to a single NUMA (Non-Uniform Memory Access) domain. This paper conducts a systematical study on the performance of multi-container deployments with different network fabrics and protocols, focusing especially on Infiniband networks. We analyze the impact of container granularity and its potential to exploit processor and memory affinity to improve applications’ performance. Our results show that default Singularity can achieve near bare-metal performance but does not support fine-grain multi-container deployments. Docker and Singularity-instance have similar behavior in terms of the performance of deployment schemes with different container granularity and affinity. This behavior differs for the several network fabrics and protocols, and depends as well on the application communication patterns and the message size. Moreover, deployments on Infiniband are also more impacted by the computation and memory allocation, and because of that, they can exploit the affinity better., We thank Lenovo for providing the testbed to run the experiments in this paper. This work was partially supported by Lenovo as part of Lenovo-BSC collaboration agreement, by the Spanish Government under contract PID2019-107255GB-C22, and by the Generalitat de Catalunya under contract 2017-SGR-1414 and under Grant No. 2020 FI-B 00257., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Heterogeneous programming using OpenMP and CUDA/HIP for hybrid CPU-GPU scientific applications

  • González Tallada, Marc
  • Morancho Llena, Enrique
Hybrid computer systems combine compute units (CUs) of different nature like CPUs, GPUs and FPGAs. Simultaneously exploiting the computing power of these CUs requires a careful decomposition of the applications into balanced parallel tasks according to both the performance of each CU type and the communication costs among them. This paper describes the design and implementation of runtime support for OpenMP hybrid GPU-CPU applications, when mixed with GPU-oriented programming models (e.g. CUDA/HIP). The paper describes the case for a hybrid multi-level parallelization of the NPB-MZ benchmark suite. The implementation exploits both coarse-grain and fine-grain parallelism, mapped to compute units of different nature (GPUs and CPUs). The paper describes the implementation of runtime support to bridge OpenMP and HIP, introducing the abstractions of Computing Unit and Data Placement. We compare hybrid and non-hybrid executions under state-of-the-art schedulers for OpenMP: static and dynamic task schedulings. Then, we improve the set of schedulers with two additional variants: a memorizing-dynamic task scheduling and a profile-based static task scheduling. On a computing node composed of one AMD EPYC 7742 @ 2.250 GHz (64 cores and 2 threads/core, totalling 128 threads per node) and 2 × GPU AMD Radeon Instinct MI50 with 32 GB, hybrid executions present speedups from 1.10× up to 3.5× with respect to a non-hybrid GPU implementation, depending on the number of activated CUs., This work was supported by the Spanish Ministry of Science and Technology (PID2019-107255GB)., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Temporally coherent video cartoonization for animation scenery generation

  • Rayo Hernández, Gustavo Enrique
  • Tous Liesa, Rubén
The automatic transformation of short background videos from real scenarios into other forms with a visually pleasing style, like those used in cartoons, holds application in various domains. These include animated films, video games, advertisements, and many other areas that involve visual content creation. A method or tool that can perform this task would inspire, facilitate, and streamline the work of artists and people who produce this type of content. This work proposes a method that integrates multiple components to translate short background videos into other forms that contain a particular style. We apply a fine-tuned latent diffusion model with an image-to-image setting, conditioned with the image edges (computed with holistically nested edge detection) and CLIP-generated prompts to translate the keyframes from a source video, ensuring content preservation. To maintain temporal coherence, the keyframes are translated into grids and the style is interpolated with an example-based style propagation algorithm. We quantitatively assess the content preservation and temporal coherence using CLIP-based metrics over a new dataset of 20 videos translated into three distinct styles., This work is partially supported by the Spanish Ministry of Science and Innovation under contract PID2019-107255GB, and by the SGR programme 2021-SGR-00478 of the Catalan Government., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Cache-aware optimization of matrix multiplication and matrix factorizations on multicore processors

  • Martínez Pérez, Héctor
  • Catalán Pallarés, Sandra
  • Igual Peña, Francisco D.
  • Herrero Zaragoza, José Ramón
  • Rodríguez Sánchez, Rafael
  • Quintana Ortí, Enrique Salvador
This paper advocates for a careful customization of the special general matrix multiplication (GEMM) kernels that are invoked from blocked routines for several relevant matrix factorizations in LAPACK, in order to improve their performance on modern multicore processors with hierarchical cache memories. To achieve this, we leverage a refined analytical model to dynamically tune the cache configuration parameters of GEMM for these kernels, taking into account the matrix operands’ dimensions, in order to improve cache occupation. In addition, toward the same goal, we accommodate a flexible development of architecture-specific micro-kernels for GEMM that allows us to select the option that, depending on the operands’ dimensions, ameliorates cache utilization. Our experiments for the LU and QR factorizations on two platforms, equipped with ARM (NVIDIA Carmel) and x86 (AMD EPYC) multi-core processors, demonstrate the benefits of this approach in terms of a better cache utilization and, in general, higher performance. Moreover, they also reveal the delicate balance between optimizing for multi-threaded parallelism versus cache usage as well as the positive effects of software prefetching., This work was supported by grants PID2020- 113656RB-C22, PID2019-107255GB, PID2021-126576NB-I00 and PID2021-123627OB-C52 of MCIN/AEI/10.13039/501100011033, by ‘‘ERDF A way of making Europe’’, and 2021-SGR-01007 of the Generalitat de Catalunya. Héctor Martínez is a postdoctoral fellow supported by the Consejería de Transformación Económica, Industria, Conocimiento y Universidades de la Junta de Andalucía. Sandra Catalán was supported by the grant RYC2021-033973- I, funded by MCIN/AEI/10.13039/501100011033 and the European Union ‘‘NextGenerationEU’’/PRTR. Funding Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Compute units in OpenMP: extensions for heterogeneous parallel programming

  • González Tallada, Marc
  • Morancho Llena, Enrique
This article evaluates the current support for heterogeneous OpenMP 5.2 applications regarding the simultaneous activation of host and device computing units (e.g., CPUs, GPUs, or FPGAs). The article identifies limitations in the current OpenMP specification and describes the design and implementation of novel OpenMP extensions and runtime support for heterogeneous parallel programming. The Compute Unit (CUs) abstraction is introduced in the OpenMP programming model. The Compute Unit abstraction is defined in terms of an aggregation of computing elements (e.g., CPUs, GPUs, FPGAs). On top of CUs, the article describes dynamic work sharing constructs and schedulers that address the inherent differences in compute power of host and device CUs. New constructs and the corresponding runtime support are described for the new abstractions. The article evaluates the case of a hybrid multilevel parallelization of the NPB-MZ benchmark suite. The implementation exploits both coarse-grain and fine-grain parallelism, mapped to CUs of different nature (GPUs and CPUs). All CUs are activated using the new extensions and runtime support. We compare hybrid and nonhybrid executions under two state-of-the-art work-distribution schemes (Static and Dynamic Task schedulers). On a computing node composed of one AMD EPYC 7742 @ 2.250GHz (64 cores and 2 threads/core, totalling 128 threads per node) and 2 GPU AMD Radeon Instinct MI50 with 32GB, hybrid executions present speedups from 1.08 up to 3.18 with respect to a nonhybrid GPU implementation, depending on the number of activated CUs., This work was supported by the Spanish Ministry of Science and Technology (PID2019-107255GB)., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Performance comparison of multi-container deployment schemes for HPC workloads: an empirical study

  • Liu, Peini
  • Guitart Fernández, Jordi
The high-performance computing (HPC) community has recently started to use containerization to obtain fast, customized, portable, flexible, and reproducible deployments of their workloads. Previous work showed that deploying an HPC workload into a single container can keep bare-metal performance. However, there is a lack of research on multi-container deployments that partition the processes belonging to each application into different containers. Partitioning HPC applications has shown to improve their performance on virtual machines by allowing to set affinity to a non-uniform memory access (NUMA) domain for each of them. Consequently, it is essential to understand the performance implications of distinct multi-container deployment schemes for HPC workloads, focusing on the impact of the container granularity and its combination with processor and memory affinity. This paper presents a systematic performance comparison and analysis of multi-container deployment schemes for HPC workloads on a single-node platform, which considers different containerization technologies (including Docker and Singularity), two different platform architectures (UMA and NUMA), and two application subscription modes (exact subscription and over-subscription). Our results indicate that finer-grained multi-container deployments, on the one side, can benefit the performance of some applications with low interprocess communication, especially in over-subscribed scenarios and when combined with affinity, but, on the other side, they can incur some performance degradation for communication-intensive applications when using containerization technologies that deploy isolated network namespaces., This work was partially supported by Lenovo as part of Lenovo-BSC collaboration agreement, by the Spanish Government under contract PID2019-107255GB-C22, and by the Generalitat de Catalunya under contract 2017-SGR-1414 and under grant 2020 FI-B 00257., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

NagareDB: A resource-efficient document-oriented time-series database

  • García Calatrava, Carlos
  • Becerra Fontal, Yolanda
  • Cucchietti, Fernando
  • Diví Cuesta, Carla
The recent great technological advance has led to a broad proliferation of Monitoring Infrastructures, which typically keep track of specific assets along time, ranging from factory machinery, device location, or even people. Gathering this data has become crucial for a wide number of applications, like exploration dashboards or Machine Learning techniques, such as Anomaly Detection. Time-Series Databases, designed to handle these data, grew in popularity, becoming the fastest-growing database type from 2019. In consequence, keeping track and mastering those rapidly evolving technologies became increasingly difficult. This paper introduces the holistic design approach followed for building NagareDB, a Time-Series database built on top of MongoDB—the most popular NoSQL Database, typically discouraged in the Time-Series scenario. The goal of NagareDB is to ease the access to three of the essential resources needed to building time-dependent systems: Hardware, since it is able to work in commodity machines; Software, as it is built on top of an open-source solution; and Expert Personnel, as its foundation database is considered the most popular NoSQL DB, lowering its learning curve. Concretely, NagareDB is able to outperform MongoDB recommended implementation up to 4.7 times, when retrieving data, while also offering a stream-ingestion up to 35% faster than InfluxDB, the most popular Time-Series database. Moreover, by relaxing some requirements, NagareDB is able to reduce the disk space usage up to 40%., This research was partly supported by the H2020 IoTwins project (Distributed Digital Twins for industrial SMEs: a big-data platform) funded by the EU under the call ICT-11-2018-2019, Grant Agreement No 857191, by the Spanish Ministry of Science and Innovation (contract PID2019- 107255GB) and by the Generalitat de Catalunya (contract 2017-SGR-1414)., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

HRM: merging hardware event monitors for improved timing analysis of complex MPSoCs

  • Vilardell Moreno, Sergi
  • Serra Mochales, Isabel
  • Santalla, Roberto
  • Mezzetti, Enrico
  • Abella Ferrer, Jaume
  • Cazorla Almeida, Francisco Javier
The Performance Monitoring Unit (PMU) in MPSoCs is at the heart of the latest measurement-based timing analysis techniques in Critical Embedded Systems. In particular, hardware event monitors (HEMs) in the PMU are used as building blocks in the process of budgeting and verifying software timing by tracking and controlling access counts to shared resources. While the number of HEMs in current MPSoCs reaches hundreds, they are read via Performance Monitoring Counters whose number is usually limited to 4-8, thus requiring multiple runs of each experiment in order to collect all desired HEMs. Despite the effort of engineers in controlling the execution conditions of each experiment, the complexity of current MPSoCs makes it arguably impossible to completely remove the noise affecting each run. As a result, HEMs read in different runs are subject to different variability, and hence, those HEMs captured in different runs cannot be ‘blindly’ merged. In this work, we focus on the NXP T2080 platform where we observed up to 59% variability across different runs of the same experiment for some relevant HEMs (e.g. processor cycles). We develop a HEM reading and merging (HRM) approach to join reliably HEMs across different runs as a fundamental element of any measurement-based timing budgeting and verification technique. Our method builds on order statistics and the selection of an anchor HEM read in all runs to derive the most plausible combination of HEM readings that keep the distribution of each HEM and their relationship with the anchor HEM intact., This work has been partially supported by the Spanish Ministry of Science and Innovation under grant PID2019-107255GB, 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., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

OmpSs@FPGA framework for high performance FPGA computing

  • De Haro Ruiz, Juan Miguel
  • Bosch Pons, Jaume
  • Filgueras Izquierdo, Antonio
  • Vidal, Miquel
  • Jiménez González, Daniel
  • Álvarez Martínez, Carlos
  • Martorell Bofill, Xavier
  • Ayguadé Parra, Eduard
  • Labarta Mancho, Jesús José
This paper presents the new features of the OmpSs@FPGA framework. OmpSs is a data-flow programming model that supports task nesting and dependencies to target asynchronous parallelism and heterogeneity. OmpSs@FPGA is the extension of the programming model addressed specifically to FPGAs. OmpSs environment is built on top of Mercurium source to source compiler and Nanos++ runtime system. To address FPGA specifics Mercurium compiler implements several FPGA related features as local variable caching, wide memory accesses or accelerator replication. In addition, part of the Nanos++ runtime has been ported to hardware. Driven by the compiler this new hardware runtime adds new features to FPGA codes, such as task creation and dependence management, providing both performance increases and ease of programming. To demonstrate these new capabilities, different high performance benchmarks have been evaluated over different FPGA platforms using the OmpSs programming model. The results demonstrate that programs that use the OmpSs programming model achieve very competitive performance with low to moderate porting effort compared to other FPGA implementations., This work has received funding from EuroEXA project (European Union’s Horizon 2020 Research and Innovation Programme, under grant agreement No 754337), from Spanish Government (projects PID2019-107255GB and SEV-2015- 0493, grant BES-2016-078046), and Generalitat de Catalunya (contracts 2017-SGR-1414 and 2017-SGR-1328)., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Embedded scaffolding for teaching and assessing inquiry-based hands-on laboratory on distributed systems

  • Guitart Fernández, Jordi
Context Information Technology education must cultivate proficiency on distributed systems, including strong hands-on laboratory skills, to meet the needs of the society and the industry. Given the complexity of distributed systems, any successful methodology to teach them to novice students must be scaffolded appropriately to ensure that the students acquire the required degree of expertise. Objective We propose a comprehensive scaffolding approach for inquiry-based hands-on laboratory on a distributed systems course, which guides not only the learning process, but also its assessment. The approach is based mainly on embedded scaffolds, namely explicit coding and experimental milestones and open questions with predefined grades, but also features contingent scaffolds provided by the teacher when additional assistance is needed. Method We apply the methodology in the context of the subject ‘Distributed Network Systems’ offered by our university. We compare the students' performance during three academic courses using the proposed methodology with respect to the three previous courses that were still using the former methodology. We use both visual representations and planned Analysis of Variance (ANOVA) tests to verify our hypothesis defined as a complex contrast. Findings We find that there is a statistically significant improvement in the students' performance when using the new methodology, both in their grades of the assignments (F(1, 75.364) = 17.770, p = 6.85*10^-5) and, more importantly, also in their grades of the exam questions about the practicals (F(1, 123.186) = 13.285, p = 3.93*10^-4). Implications Our results encourage other instructors to incorporate embedded scaffolds for teaching and assessing their hands-on laboratories on distributed systems., This research was partially supported by the Agencia Estatal de Investigación under contracts PID2019-107255GB-C22 and PID2021-126248OB-I00 and by the Generalitat de Catalunya under contract 2021-SGR-00478., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Near-optimal replacement policies for shared caches in multicore processors

  • Díaz Maag, Javier
  • Ibáñez Marín, Pablo
  • Monreal Arnal, Teresa
  • Viñals Yúfera, Victor
  • Llaberia Griñó, José M.
An optimal replacement policy that minimizes the miss rate in a private cache was proposed several decades ago. It requires knowing the future access sequence the cache will receive. There is no equivalent for shared caches because replacement decisions alter this future sequence. We present a novel near-optimal policy for minimizing the miss rate in a shared cache that approaches the optimal execution iteratively. During each iteration, the future access sequence is reconstructed on every miss interleaving the future per-core sequences, taken from the previous iteration. This single sequence feeds a classical private-cache optimum replacement policy. Our evaluation on a shared last-level cache shows that our proposal iteratively converges to a near-optimal miss rate that is independent of the initial conditions, within a margin of 0.1%. The best state-of-the-art online policies achieve around 65% of the miss rate reduction obtained by our near-optimal proposal. In a shared cache, miss rate optimization does not imply the optimization of other metrics. Therefore, we also propose a new near-optimal policy to maximize fairness between cores. The best state-of-the-art online policy achieves 60% of the improvement in fairness seen with our near-optimal policy. Our proposals are useful both for setting upper performance bounds and inspiring implementable mechanisms for shared caches., The authors acknowledge support from grants (1) PID2019-105660RB-C21 and PID2019-107255GB-C22 from Agencia Estatal de Investigación (AEI) from Spain and European Regional Development Fund (ERDF); (2) gaZ: T58_20R research group from Aragón Government and European Social Fund (ESF); and (3) 2014-2020 "Construyendo Europa desde Aragón" from European Regional Development Fund (ERDF)., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Practicable live container migrations in high performance computing clouds: Diskless, iterative, and connection-persistent

  • Guitart Fernández, Jordi
Checkpoint/Restore techniques had been thoroughly used by the High Performance Computing (HPC) community in the context of failure recovery. Given the current trend in HPC to use containerization to obtain fast, customized, portable, flexible, and reproducible deployments of their workloads, as well as efficient and reliable sharing and management of HPC Cloud infrastructures, there is a need to integrate Checkpoint/Restore with containerization in such a way that the freeze time of the application is minimal and live migrations are practicable. Whereas current Checkpoint/Restore tools (such as CRIU) support several options to accomplish this, most of them are rarely exploited in HPC Clouds and, consequently, their potential impact on the performance is barely known. Therefore, this paper explores the use of CRIU’s advanced features to implement diskless, iterative (pre-copy and post-copy) migrations of containers with external network namespaces and established TCP connections, so that memory-intensive and connection-persistent HPC applications can live-migrate. Our extensive experiments to characterize the performance impact of those features demonstrate that properly-configured live migrations incur low application downtime and memory/disk usage and are indeed feasible in containerized HPC Clouds., This research was partially supported by the Spanish Government under contract PID2019-107255GB-C22, by the Generalitat de Catalunya, Spain under contract 2021-SGR-00478, and by the EU-HORIZON programme under grant agreement 101092646., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Acceleration strategies for large-scale sequential simulations using parallel neighbour search: Non-LVA and LVA scenarios

  • Peredo Andrade, Oscar Francisco
  • Herrero Zaragoza, José Ramón
This paper describes the application of acceleration techniques into existing implementations of Sequential Gaussian Simulation and Sequential Indicator Simulation. These implementations might incorporate Locally Varying Anisotropy (LVA) to capture non-linear features of the underlying physical phenomena. The imple- mentation focuses on a novel parallel neighbour search algorithm, which can be used on both non-LVA and LVA codes. Additionally, parallel shortest path executions and optimized linear algebra libraries are applied with focus on LVA codes. Execution time, speedup and accuracy results are presented. Non-LVA codes are benchmarked using two scenarios with approximately 50 million domain points each. Speedup results of 2× and 4× were obtained on SGS and SISIM respectively, where each scenario is compared against a baseline code published in Peredo et al. (2018). The aggregated contribution to speedup of both works results in 12× and 50× respectively. LVA codes are benchmarked using two scenarios with approximately 1.7 million domain points each. Speedup results of 56× and 1822× were obtained on SGS and SISIM respectively, where each scenario is compared against the original baseline sequential codes., The authors acknowledge the donated resources from project PID2019-107255GB of the Spanish Ministerio de Economía y Competitividad, and project 2017-SGR-1414 from the Generalitat de Catalunya, Spain., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Exploring the potential of BCI in education: an experiment in musical training

  • Folgieri, Raffaella
  • Lucchiari, Claudio
  • Gricar, Sergej
  • Baldigara, Tea
  • Gil, Marisa
Brain–computer interfaces (BCIs) have gained significant attention in recent years for various applications, including education and skill development: studies have shown that BCIs can boost memory, concentration, and even creativity and can improve learning and memory retention in healthy people. In our current study, we investigated the effectiveness of real-time feedback provided by a BCI system for improving performance on a specific task. A total of 20 participants completed a pre-training assessment, followed by a training period with the BCI system and a post-training assessment. The BCI system provided real-time feedback based on the participants’ level of accuracy, with positive feedback given for scores above 70%. Results showed a significant improvement in accuracy scores from pre- to post-training, with an average improvement of 15%. Participants also reported high levels of satisfaction with the feedback provided by the BCI system. These findings suggest that real-time feedback provided by a BCI system can be an effective tool for skill development and education, particularly when tailored to the specific needs of individual learners. Further research is needed to explore the potential of BCIs for a wide range of educational applications., This research was funded by the Department of Philosophy “Piero Martinetti” of the University of Milan under the Project “Departments of Excellence 2023–2027” awarded by the Ministry of University and Research (MUR). Marisa Gil’s work has been supported by the Spanish Ministry of Education (PID2019-107255GB-C22) and partially supported by the Generalitat de Catalunya (contract 2021-SGR-01007). Certain sections of this article were written using tools funded by the Slovenian Research and Innovation Agency under grant number CRP2023 V5-2331. The same institutions funded the APC., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Mixed-precision pre-pivoting strategy for the LU factorization

  • Sahraneshinsamani, Nima
  • Catalán Pallarés, Sandra
  • Herrero Zaragoza, José Ramón
This paper investigates the efficient application of half-precision floating-point (FP16) arithmetic on GPUs for boosting LU decompositions in double (FP64) precision. Addressing the motivation to enhance computational efficiency, we introduce two novel algorithms: Pre-Pivoted LU (PRP) and Mixed-precision Panel Factorization (MPF). Deployed in both hybrid CPU-GPU setups and native GPU-only configurations, PRP identifies pivot lists through LU decomposition computed in reduced precision and subsequently reorders matrix rows in FP64 precision before executing LU decomposition without pivoting. Two variants of PRP, namely hPRP and xPRP, are introduced, differing in their computation of pivot lists in full half-precision or mixed half-single precision. The MPF algorithm generates FP64 LU factorization while internally utilizing hPRP for panel factorization, showcasing accuracy on par with standard DGETRF but with superior speed. The study further explores auxiliary functions required for the native mode implementation of PRP variants and MPF., The authors acknowledge the funding received from project PID2019-107255GB of the Spanish Ministerio de Economía Competitividad, and project 2021-SGR-01007 from the Generalitat de Catalunya, Spain. Sandra Catalán was supported by the grant RYC2021-033973-I, funded by MCIN/AEI/10.13039/501100011033 and the European Union “NextGenerationEU”/PRTR, and UJI-2023-04, funded by Universitat Jaume I., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

A differential privacy protection-based federated deep learning framework to fog-embedded architectures

  • Gutiérrez Escobar, Norma
  • Otero Calviño, Beatriz
  • Rodríguez Luna, Eva
  • Utrera Iglesias, Gladys Miriam
  • Mus León, Sergi
  • Canal Corretger, Ramon
Nowadays, companies collect massive quantities of data to enhance their operations, often at the expense of sharing user sensible information. This data is widely used to train Deep Learning (DL) neural networks to model, classify, or recognize complex data. These activities enable companies to offer an array of services to users, such as precise advertising and optimal location services. This study explores potential solutions for preserving privacy while utilizing DL applications. To address the privacy issue, we develop a privacy-preserving framework specifically designed for fog computing environments. Unlike traditional cloud computing architectures, fog embedded architectures only share a small portion of user data with a nearby fog node, ensuring that the majority of sensitive data remains secure. Within these fog nodes, we incorporate two additional algorithms, namely Generalization and Threshold, to enhance the privacy-preserving capabilities of the framework. The first algorithm, Generalization, introduces a validation dataset within the fog nodes which not only increases the accuracy of the fog-embedded framework but also ensures that user data is preserved. The second algorithm, Threshold, is responsible for protecting user data samples and reducing the amount of information sent to the server. By combining these two algorithms, we are able to provide an additional layer of protection for user privacy while still maintaining the accuracy of the model. We conduct an evaluation to test its effectiveness using two separate datasets. In addition, we analyze them through a Feed Forward Neural Network (FFNN) and compare the results with a traditional centralized architecture to validate the effectiveness of the proposed framework. The results of our evaluation demonstrate that the proposed privacy-preserving framework, when combined with the Generalization and Threshold algorithms, can preserve up to 38.44% of user data. Additionally, we were able to extend the framework to multiple fog nodes without compromising the network’s accuracy, as we only observed a 0.1% decrease in accuracy when using the proposed architecture. This study emphasizes the importance of preserving user information while using DL applications and provides a solution that trains the desired network without violating user privacy, hence preserving their anonymity. Overall, the study highlights the potential of Federated Deep Learning to improve the accuracy and privacy of DL applications in fog computing environments., This work is partially supported by the Spanish Ministry of Science and Innovation, Spain under contracts PID2021-124463OB-IOO and PID2019-107255GB-C22; by the Generalitat de Catalunya, Spain under grant 2021SGR00326; and by the DRAC (IU16-011591), the HORIZON Vitamin-V, Spain(101093062), the HORIZON-AG PHOENI2X, Spain (101070586) and the HORIZON HORSE, Spain (101096342) projects., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Open project-based learning for dynamic adaptability of IT education

  • Tous Liesa, Rubén
  • Freitag, Fèlix
The pace of technological change is accelerating, the duration of the trends is becoming shorter and the number of technological alternatives to solve IT problems grows quickly. Keeping IT higher education course curricula up-to-date becomes more and more challenging. This paper presents an empirical study that aims to assess whether an open-statement and open-solution methodology in an IT-related project-based learning (PBL) course led to the adaptation of the skills learned by students to the trends of the main IT technology domains. The study analyses data from more than 90 projects by students from eight academic years of the course” Project on Information Technologies”, carried out by students in the last year of the Bachelor of Science in Informatics Engineering at the Barcelona School of Informatics. Our results identify a high correlation between the technologies chosen in student projects and technological trends, which shows that the methodology encourages students to incorporate the latest technological innovations in their project work plan., This work is partially supported by the Spanish Ministry of Economy and Competitivity under contract TIN2015-65316-P, by the Spanish Ministry of Science and Innovation under contracts PID2019-107255GB and PID2019-106774RB-C21, and by the SGR programmes (2017-SGR-962 and 2017-SGR-990) of the Catalan Government., Peer Reviewed, Postprint (author's final draft)