Pasar al contenido principal

Resultados relacionados

Resultados totales (Incluyendo duplicados): 48
Encontrada(s) 1 página(s)

DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

On the maturity of parallel applications for asymmetric multi-core processors

  • Chronaki, Kallia
  • Moreto, Miquel
  • Casas, Marc
  • Rico, Alejandro
  • Badia, Rosa M.
  • Ayguade, Eduard
  • Valero, Mateo
Asymmetric multi-cores (AMCs) are a successful architectural solution for both mobile devices and supercomputers. By maintaining two types of cores (fast and slow) AMCs are able to provide high performance under the facility power budget. This paper performs the first extensive evaluation of how portable are the current HPC applications for such supercomputing systems. Specifically we evaluate several execution models on an ARM big.LITTLE AMC using the PARSEC benchmark suite that includes representative highly parallel applications. We compare schedulers at the user, OS and runtime levels, using both static and dynamic options and multiple configurations, and assess the impact of these options on the well-known problem of balancing the load across AMCs. Our results demonstrate that scheduling is more effective when it takes place in the runtime system level as it improves the baseline by 23%, while the heterogeneous-aware OS scheduling solution improves the baseline by 10%., This work has been supported by the RoMoL ERC Advanced Grant (GA 321253), by the European HiPEAC Network of Excellence, by the Spanish Ministry of Science and Innovation (contracts TIN2015-65316-P), by the Generalitat de Catalunya (contracts 2014-SGR-1051 and 2014-SGR-1272), and by the European Union’s Horizon 2020 research and innovation programme under grant agreement No 671697 and No. 779877. Kallia Chronaki has been partially supported by the Ministry of Economy and Competitiveness under Ramon y Cajal fellowship number RYC-2016-21104., Peer reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

On the maturity of parallel applications for asymmetric multi-core processors

  • Chronaki, Kallia
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Casas, Marc|||0000-0003-4564-2093
  • Rico, Alejandro
  • Badia Sala, Rosa Maria|||0000-0003-2941-5499
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
  • Valero Cortés, Mateo|||0000-0003-2917-2482
Asymmetric multi-cores (AMCs) are a successful architectural solution for both mobile devices and supercomputers. By maintaining two types of cores (fast and slow) AMCs are able to provide high performance under the facility power budget. This paper performs the first extensive evaluation of how portable are the current HPC applications for such supercomputing systems. Specifically we evaluate several execution models on an ARM big.LITTLE AMC using the PARSEC benchmark suite that includes representative highly parallel applications. We compare schedulers at the user, OS and runtime levels, using both static and dynamic options and multiple configurations, and assess the impact of these options on the well-known problem of balancing the load across AMCs. Our results demonstrate that scheduling is more effective when it takes place in the runtime system level as it improves the baseline by 23%, while the heterogeneous-aware OS scheduling solution improves the baseline by 10%., This work has been supported by the RoMoL ERC Advanced Grant (GA 321253), by the European HiPEAC Network of Excellence, by the Spanish Ministry of Science and Innovation (contracts TIN2015-65316-P), by the Generalitat de Catalunya (contracts 2014-SGR-1051 and 2014-SGR-1272), and by the European Union's Horizon 2020 research and innovation programme under grant agreement No 671697 and No. 779877. M. Moretó has been partially supported by the Ministry of Economy and Competitiveness under Ramon y Cajal fellowship number RYC-2016-21104., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Runtime-assisted cache coherence deactivation in task parallel programs

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

Design space exploration of next-generation HPC machines

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

TaskGenX: A Hardware-Software Proposal for Accelerating Task Parallelism

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

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

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

On the benefits of tasking with OpenMP

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

Demystifying the Characteristics of High Bandwidth Memory for Real-Time Systems

  • Asifuzzaman, Kazi
  • Abuelala, Mohamed
  • Hassan, Mohamed
  • Cazorla Almeida, Francisco Javier
The number of functionalities controlled by software on every critical real-time product is on the rise in domains like automotive, avionics and space. To implement these advanced functionalities, software applications increasingly adopt artificial intelligence algorithms that manage massive amounts of data transmitted from various sensors. This translates into unprecedented memory performance requirements in critical systems that the commonly used DRAM memories struggle to provide. High-Bandwidth Memory (HBM) can satisfy these requirements offering high bandwidth, low power and high-integration capacity features. However, it remains unclear whether the predictability and isolation properties of HBM are compatible with the requirements of critical embedded systems. In this work, we perform to our knowledge the first timing analysis of HBM. We show the unique structural and timing characteristics of HBM with respect to DRAM memories and how they can be exploited for better time predictability, with emphasis on increased isolation among tasks and reduced worst-case memory latency., This work has been partially supported by the Spanish Ministry of Science and Innovation under grant PID2019-107255GB-C21/AEI/10.13039/501100011033; the European Union’s Horizon 2020 Framework Programme under grant agreement No. 878752 (MASTECS) and agreement No. 779877 (Mont-Blanc 2020); the European Research Council (ERC) grant agreement No. 772773 (SuPerCom); and the Natural Sciences and Engineering Research Council of Canada (NSERC), Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Improving predication efficiency through compaction/restoration of SIMD instructions

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

PrioRAT: criticality-driven prioritization inside the on-chip memory hierarchy

  • Dimic, Vladimir
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Casas, Marc|||0000-0003-4564-2093
  • Valero Cortés, Mateo|||0000-0003-2917-2482
The ever-increasing gap between the processor and main memory speeds requires careful utilization of the limited memory link. This is additionally emphasized for the case of memory-bound applications. Prioritization of memory requests in the memory controller is one of the approaches to improve performance of such codes. However, current designs do not consider high-level information about parallel applications. In this paper, we propose a holistic approach to this problem, where the runtime system-level knowledge is made available in hardware. Processor exploits this information to better prioritize memory requests, while introducing negligible hardware cost. Our design is based on the notion of critical path in the execution of a parallel code. The critical tasks are accelerated by prioritizing their memory requests within the on-chip memory hierarchy. As a result, we reduce the critical path and improve the overall performance up to 1.19× compared to the baseline systems., This work has been partially supported by the Spanish Ministry of Science and Innovation (PID2019-107255GB-C21/AEI/10.13039/ 501100011033), by the Generalitat de Catalunya (contracts 2017-SGR-1414 and 2017-SGR-1328), by the European Unions Horizon 2020 research and innovation program under the Mont-Blanc 2020 project (grant agreement 779877) and by the RoMoL ERC Advanced Grant (GA 321253). V. Dimić has been partially supported by the Agency for Management of University and Research Grants (AGAUR) of the Government of Catalonia under Ajuts per a la contractaci o de personal investigador novell fellowship number 2017 FI B 00855. M. Moreto and M. Casas have been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramon y Cajal fellowship numbers RYC- 2016-21104 and RYC-2017-23269, respectively., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Wavefront parallelization of recurrent neural networks on multi-core architectures

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

Teaching HPC systems and parallel programming with small-scale clusters

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

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

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

Design trade-offs for emerging HPC processors based on mobile market technology

  • Armejach Sanosa, Adrià|||0000-0003-2869-668X
  • Casas, Marc|||0000-0003-4564-2093
  • Moretó Planas, Miquel|||0000-0002-9848-8758
This is a post-peer-review, pre-copyedit version of an article published in The Journal of Supercomputing. The final authenticated version is available online at: http://dx.doi.org/10.1007/s11227-019-02819-4, High-performance computing (HPC) is at the crossroads of a potential transition toward mobile market processor technology. Unlike in prior transitions, numerous hardware vendors and integrators will have access to state-of-the-art processor designs due to Arm’s licensing business model. This fact gives them greater flexibility to implement custom HPC-specific designs. In this paper, we undertake a study to quantify the different energy-performance trade-offs when architecting a processor based on mobile market technology. Through detailed simulations over a representative set of benchmarks, our results show that: (i) a modest amount of last-level cache per core is sufficient, leading to significant power and area savings; (ii) in-order cores offer favorable trade-offs when compared to out-of-order cores for a wide range of benchmarks; and (iii) heterogeneous configurations help to improve processor performance and energy efficiency., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

gem5 + rtl: A framework to enable RTL models inside a full-system simulator

  • López Paradís, Guillem
  • Armejach Sanosa, Adrià|||0000-0003-2869-668X
  • Moretó Planas, Miquel|||0000-0002-9848-8758
In recent years there has been a surge of interest in designing custom accelerators for power-efficient high-performance computing. However, available tools to simulate low-level RTL designs often neglect the target system in which the design will operate. This hinders proper testing and debugging of functionalities, and does not allow co-designing the accelerator to obtain a balanced and efficient architecture. In this paper, we introduce gem5 + rtl, a flexible framework that enables simulation of RTL models inside a full-system software simulator. We present the framework’s functionality that allows easy integration of RTL models on a simulated system-on-chip (SoC) that is able to boot Linux and run complex multi-threaded and multi-programmed workloads. We demonstrate the framework with two relevant use cases that integrate a multi-core SoC with a Performance Monitoring Unit (PMU) and the NVIDIA Deep Learning Accelerator (NVDLA), showcasing how the framework enables testing RTL model features and how it can enable co-design taking into account the entire SoC., This research was supported by the European Union Regional Development Fund within the framework of the ERDF Operational Program of Catalonia 2014-2020 with a grant of 50% of total cost eligible under the DRAC project [001-P- 001723], by the Spanish goverment (grant RTI2018-095094- B-C21 CONSENT), by the Spanish Ministry of Science and Innovation (contracts PID2019-107255GB-C21) and by the Catalan Government (contracts 2017-SGR-1414, 2017-SGR705). This work has also been supported by the European Community’s Horizon 2020 Framework Programme under the Mont-Blanc 2020 and EPI projects (grant agreements n. 779877 and n. 826647); and by the Arm-BSC Center of Excellence. G. López-Paradís has been partially supported by the Agency for Management of University and Research Grants (AGAUR) of the Government of Catalonia under Ajuts per a la contractació de personal investigador novell fellowship No. 2021FI B00994. A. Armejach has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Juan de la Cierva postdoctoral fellowship number IJCI-2017-33945. M. Moretó has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramón y Cajal fellowship No. RYC-2016-21104., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Stencil codes on a vector length agnostic architecture

  • Armejach Sanosa, Adrià|||0000-0003-2869-668X
  • Caminal Pallarés, Helena
  • Cebrián González, Juan Manuel
  • González-Alberquilla, Rekai
  • Adeniyi-Jones, Chris
  • Valero Cortés, Mateo|||0000-0003-2917-2482
  • Casas, Marc|||0000-0003-4564-2093
  • Moretó Planas, Miquel|||0000-0002-9848-8758
Data-level parallelism is frequently ignored or underutilized. Achieved through vector/SIMD capabilities, it can provide substantial performance improvements on top of widely used techniques such as thread-level parallelism. However, manual vectorization is a tedious and costly process that needs to be repeated for each specific instruction set or register size. In addition, automatic compiler vectorization is susceptible to code complexity, and usually limited due to data and control dependencies. To address some these issues, Arm recently released a new vector ISA, the Scalable Vector Extension (SVE), which is Vector-Length Agnostic (VLA). VLA enables the generation of binary files that run regardless of the physical vector register length. In this paper we leverage the main characteristics of SVE to implement and optimize stencil computations, ubiquitous in scientific computing. We show that SVE enables easy deployment of textbook optimizations like loop unrolling, loop fusion, load trading or data reuse. Our detailed simulations using vector lengths ranging from 128 to 2,048 bits show that these optimizations can lead to performance improvements over straight-forward vectorized code of up to 56.6% for 2,048 bit vectors. In addition, we show that certain optimizations can hurt performance due to a reduction in arithmetic intensity, and provide insight useful for compiler optimizers., This work has been partially supported by the European HiPEAC Network of Excellence, by the Spanish Ministry of Economy and Competitiveness (contract TIN2015-65316-P), and by the Generalitat de Catalunya (contracts 2017-SGR-1328 and 2017-SGR-1414). The Mont-Blanc project receives funding from the EUs H2020 Framework Programme (H2020/2014-2020) under grant agreements no. 671697 and no. 779877. M. Moreto has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramon y Cajal fellowship number RYC-2016-21104. Finally, A. Armejach has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Juan de la Cierva postdoctoral fellowship number FJCI-2015-24753., 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)

Semi-automatic validation of cycle-accurate simulation infrastructures: The case for gem5-x86

  • Cebrián González, Juan Manuel
  • Barredo Ferreira, Adrián
  • Caminal Pallarés, Helena
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Casas, Marc|||0000-0003-4564-2093
  • Valero Cortés, Mateo|||0000-0003-2917-2482
Since the early 70s, simulation infrastructures have been a keystone in computer architecture research, providing a fast and reliable way to prototype and evaluate ideas for future computing systems. There are different types of simulators, from most detailed (cycle-accurate) to time-based/functional and analytical modeling. Increasing accuracy translates into several orders of magnitude in terms of simulation speed. Yet, a question remains open: are the results derived from the simulation infrastructure representative of a real machine? Validation of these infrastructures is complex and costly, usually performed upon release. However, most simulators do not provide the appropriate means to verify or validate new architectural models. In this paper, we introduce a semi-automatic validation framework based on real-hardware performance counter information. The framework provides two levels of abstraction: (a) a high level definition of the processor behavior (Top-Down model) and (b) detailed per-structure and per-pipeline-stage usage breakdown to pinpoint simulator issues. We used this framework to validate the latest available gem5-x86 simulation environment, and found several sources of error that alter the expected behavior of the simulated processor, which we were later to document and correct., This work has been partially supported by the Spanish Government (Severo Ochoa grants SEV2015-0493, SEV-2011-00067), the Spanish Ministry of Science and Innovation (contract TIN2015- 65316-P), Generalitat de Catalunya, Spain (contracts 2014-SGR1051 and 2014-SGR-1272), the RoMoL ERC Advanced, Spain Grant (GA 321253), the European HiPEAC Network of Excellence and the Mont-Blanc project (EU-FP7-610402 and EU-H2020-779877). A. Barredo has been supported by the Spanish Government under Formación del Personal Investigador fellowship number BES2017-080635. M. Moreto and M. Casas have been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness, Spain under Ramon y Cajal fellowship numbers RYC2016-21104 and RYC-2017-23269., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Mont-Blanc 2020: Towards scalable and power efficient European HPC processors

  • Armejach Sanosa, Adrià|||0000-0003-2869-668X
  • Brank, Bine
  • Cortina Guardia, Jordi
  • Dolique, François
  • Hayes, Timothy
  • Ho, Nam
  • Lagadec, Pierre-Axel
  • Lemaire, Romain
  • López Paradís, Guillem
  • Marliac, Laurent
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Marcuello Pascual, Pedro
  • Pleiter, Dirk
  • Tan, Xubin
  • Derradji, Said
The Mont-Blanc 2020 (MB2020) project has triggered the development of the next generation industrial processor for Big Data and High Performance Computing (HPC). MB2020 is paving the way to the future low-power European processor for exascale, defining the System-on-Chip (SoC) architecture and implementing new critical building blocks to be integrated in such an SoC. In this paper, we first present an overview of the MB2020 project, then we describe our experimental infrastructure, the requirements of relevant applications, and the IP blocks developed in the project. Finally, we present our emulation-based final demonstrator and explain how it integrates within our first generation of HPC processors., This work is supported by the European Community’s Horizon 2020 Framework Programme under the Mont-Blanc 2020 project, grant agreement n. 779877., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

BST: A BookSim-based toolset to simulate NoCs with single- and multi-hop bypass

  • Pérez Gallardo, Iván
  • Vallejo Gutiérrez, Enrique
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Beivide Palacio, Julio Ramon
Network-on-Chips are a critical part of modern multiprocessors and their relevance will grow with the number of cores. The development of future NoC designs relies on detailed simulation models that accurately estimate their performance, power and hardware cost. Bypass routers are promising proposals due to their improved performance. Bypass routers reduce latency thanks to a combination of speculation, pre-routing (lookahead routing) and buffer bypass, which also reduce energy consumption by avoiding unnecessary buffer writes and reads. Multi-hop bypass NoCs, known as SMART, even bypass the crossbar of multiple routers in a single cycle. However, publicly available NoC simulators, such as BookSim or Garnet, do not implement bypass mechanisms or do not model them accurately. In this work, we present Bypass Simulation Toolset (BST), a set of tools to accurately simulate NoCs with single- and multi-hop bypass routers. BST combines and extends several simulation tools: an extension of BookSim with state-of-theart cycle-accurate bypass router models and additional flow control mechanisms; an RTL implementation of multi-hop bypass mechanisms based on OpenSMART; an API to ease a modular integration of the BST NoC simulator in full system simulators; and a set of scripts to automate simulation execution and data collection. To showcase BST, we i) validate BookSim SMART models with the RTL implementation; ii) compare bypass and traditional nonbypass router models; iii) integrate BookSim in gem5 using the proposed API and compare it with gem5’s Simple and Garnet 2.0 NoC models; and iv) present a case study evaluating different combinations of router types and topologies recently proposed for NoCs, highlighting the flexibility of the BST toolset., This work was supported by the Spanish Ministry of Science, Innovation and Universities, contracts TIN2015-65316- P, TIN2016-76635-C2-2-R (AEI/FEDER, UE) and PID2019- 105660RB-C22; the European Union’s Horizon 2020 research and innovation program under the Mont-Blanc 2020 project (grant agreement 779877); and the HiPEAC Network of Excellence. I. Perez is partially supported by an FPI grant, BES- ´ 2017-079971. M. Moreto has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramon y Cajal fellowship number RYC-2016-21104. Bluespec Inc. provided access to Bluespec tools., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Characterizing the impact of last-level cache replacement policies on big-data workloads

  • Jamet, Alexandre Valentin
  • Álvarez Martí, Lluc|||0000-0003-0506-8867
  • Jiménez, Daniel A.
  • Casas, Marc|||0000-0003-4564-2093
The vast disparity between Last Level Cache (LLC) and memory latencies has motivated the need for efficient cache management policies. The computer architecture literature abounds with work on LLC replacement policy. Although these works greatly improve over the least-recently-used (LRU) policy, they tend to focus only on the SPEC CPU 2006 benchmark suite - and more recently on the SPEC CPU 2017 benchmark suite - for evaluation. However, these workloads are representative for only a subset of current High-Performance Computing (HPC) workloads. In this paper we evaluate the behavior of a mix of graph processing, scientific and industrial workloads (GAP, XSBench and Qualcomm) along with the well-known SPEC CPU 2006 and SPEC CPU 2017 workloads on state-of-the-art LLC replacement policies such as Multiperspective Reuse Prediction (MPPPB), Glider, Hawkeye, SHiP, DRRIP and SRRIP. Our evaluation reveals that, even though current state-of-the-art LLC replacement policies provide a significant performance improvement over LRU for both SPEC CPU 2006 and SPEC CPU 2017 workloads, those policies are hardly able to capture the access patterns and yield sensible improvement on current HPC and big data workloads due to their highly complex behavior. In addition, this paper introduces two new LLC replacement policies derived from MPPPB. The first proposed replacement policy, Multi-Sampler Multiperspective (MS-MPPPB), uses multiple samplers instead of a single one and dynamically selects the best-behaving sampler to drive reuse distance predictions. The second replacement policy presented in this paper, Multiperspective with Dynamic Features Selector (DS-MPPPB), selects the best behaving features among a set of 64 features to improve the accuracy of the predictions. On a large set of workloads that stress the LLC, MS-MPPPB achieves a geometric mean speed-up of 8.3% over LRU, while DS-MPPPB outperforms LRU by a geometric mean speedup of 8.0%. For big data and HPC workloads, the two proposed techniques present higher performance benefits than state-of-the-art approaches such as MPPPB, Glider and Hawkeye, which yield geometric mean speedups of 7.0%, 5.0% and 4.8% over LRU, respectively., 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). Lluc Alvarez has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under the Juan de la Cierva Formacion fellowship number FJCI-2016-30984. Marc Casas has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramon y Cajal fellowship number RYC-2017-23269. This research was supported by National Science Foundation grant CCF-1912617, Semiconductor Research Corporation project 2936.001, and generous gifts from Intel Labs., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Data prefetching on in-order processors

  • Ortega Carrasco, Cristobal
  • García Flores, Víctor
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Casas, Marc|||0000-0003-4564-2093
  • Rositoru, Roxana
Low-power processors have attracted attention due to their energy-efficiency. A large market, such as the mobile one, relies on these processors for this very reason. Even High Performance Computing (HPC) systems are starting to consider low-power processors as a way to achieve exascale performance within 20MW, however, they must meet the right performance/Watt balance. Current low-power processors contain in-order cores, which cannot re-order instructions to avoid data dependency-induced stalls. Whilst this is useful to reduce the chip's total power consumption, it brings several challenges. Due to the evolving performance gap between memory and processor, memory is a significant bottleneck. In-order cores cannot re-order instructions and are memory latency bound, something data prefetching can help alleviate by ensuring data is readily available. In this work, we do an exhaustive analysis of available data prefetching techniques in state-of-The-Art in-order cores. We analyze 5 static prefetchers and 2 dynamic aggressiveness and destination mechanisms applied to 3 data prefetchers on a set of HPC mini-and proxy-Applications, whilst running on in-order processors. We show that next-line prefetching can achieve nearly top performance with a reasonable bandwidth consumption when throttled, whilst neighbor prefetchers have been found to be best, overall., This work has been supported by the RoMoL ERC Advanced Grant (GA 321253), by the European HiPEAC Network of Excellence, by the Spanish Ministry of Science and Innovation (contracts TIN2015-65316-P), by Generalitat de Catalunya (contracts 2014-SGR-1051 and 2014-SGR-1272), by the European Union’s Horizon 2020 research and innovation programme (grant agreements No 671697 and No 779877). M. Moreto has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramon y Cajal fellowship number RYC-2016-21104., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

A vulnerability factor for ECC-protected memory

  • Jaulmes, Luc|||0000-0002-5815-2504
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Valero Cortés, Mateo|||0000-0003-2917-2482
  • Casas, Marc|||0000-0003-4564-2093
Fault injection studies and vulnerability analyses have been used to estimate the reliability of data structures in memory. We survey these metrics and look at their adequacy to describe the data stored in ECC-protected memory. We also introduce FEA, a new metric improving on the memory derating factor by ignoring a class of false errors. We measure all metrics using simulations and compare them to the outcomes of injecting errors in real runs. This in-depth study reveals that FEA provides more accurate results than any state-of-the-art vulnerability metric. Furthermore, FEA gives an upper bound on the failure probability due to an error in memory, making this metric a tool of choice to quantify memory vulnerability. Finally, we show that ignoring these false errors reduces the failure rate on average by 12.75% and up to over 45%., This work has been supported by the RoMoL ERC Advanced Grant (GA 321253), by the European HiPEAC Network of Excellence, by the Spanish Ministry of Economy and Competitiveness (contract TIN2015-65316- P), by the Generalitat de Catalunya (contracts 2017-SGR-1414 and 2017- SGR-1328), by the Spanish Government (Severo Ochoa grant SEV-2015- 0493) and by the European Union’s Horizon 2020 research and innovation programme (grant agreements 671697 and 779877). L. Jaulmes has been partially supported by the Spanish Ministry of Education, Culture and Sports under grant FPU2013/06982. M. Moreto and M. Casas have been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramon y Cajal fellowships RYC-2016-21104 and RYC-2017-23269., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Estudio, informe, memoria (Report)

Towards resilient EU HPC systems: A blueprint

  • Radojković, Petar
  • Marazakis, Manolis
  • Carpenter, Paul Matthew
  • Jeyapaul, Reiley
  • Gizopoulos, Dimitris
  • Schulz, Martin
  • Armejach Sanosa, Adrià|||0000-0003-2869-668X
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
  • Canal Corretger, Ramon|||0000-0003-4542-204X
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Salami, Behzad
  • Unsal, Osman Sabri
This document aims to spearhead a Europe-wide discussion on HPC system resilience and to help the European HPC community define best practices for resilience. We analyse a wide range of state-of-the-art resilience mechanisms and recommend the most effective approaches to employ in large-scale HPC systems. Our guidelines will be useful in the allocation of available resources, as well as guiding researchers and research funding towards the enhancement of resilience approaches with the highest priority and utility. Although our work is focused on the needs of next generation HPC systems in Europe, the principles and evaluations are applicable globally., This work has received funding from the European Union’s Horizon 2020 research and innovation programme under the projects ECOSCALE (grant agreement No 671632), EPI (grant agreement No 826647), EuroEXA (grant agreement No 754337), Eurolab4HPC (grant agreement No 800962), EVOLVE (grant agreement No 825061), EXA2PRO (grant agreement No 801015), ExaNest (grant agreement No 671553), ExaNoDe (grant agreement No 671578), EXDCI-2 (grant agreement No 800957), LEGaTO (grant agreement No 780681), MB2020 (grant agreement No 779877), RECIPE (grant agreement No 801137) and SDK4ED (grant agreement No 780572). The work was also supported by the European Commission’s Seventh Framework Programme under the projects CLERECO (grant agreement No 611404), the NCSA-Inria-ANL-BSC-JSCRiken-UTK Joint-Laboratory for Extreme Scale Computing – JLESC (https://jlesc.github.io/), OMPI-X project (No ECP-2.3.1.17) and the Spanish Government through Severo Ochoa programme (SEV-2015-0493). This work was sponsored in part by the U.S. Department of Energy's Office of Advanced Scientific Computing Research, program managers Robinson Pino and Lucy Nowell. This manuscript has been authored by UT-Battelle, LLC under Contract No DE-AC05-00OR22725 with the U.S. Department of Energy.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publicación Artículo científico (JournalArticle)

Porting and optimizing BWA-MEM2 using the Fujitsu A64FX processor

  • Langarita, Rubén
  • Armejach, Adrià
  • Ibáñez, Pablo
  • Alastruey-Benedé, Jesús
  • Moretó, Miquel
Sequence alignment pipelines for human genomes are an emerging workload that will dominate in the precision medicine field. BWA-MEM2 is a tool widely used in the scientific community to perform read mapping studies. In this paper, we port BWA-MEM2 to the AArch64 architecture using the ARMv8-A specification, and we compare the resulting version against an Intel Skylake system both in performance and in energy-to-solution. The porting effort entails numerous code modifications, since BWA-MEM2 implements certain kernels using x86 64 specific intrinsics, e.g., AVX-512. To adapt this code we use the recently introduced Arm’s Scalable Vector Extensions (SVE). More specifically, we use Fujitsu’s A64FX processor, the first to implement SVE. The A64FX powers the Fugaku Supercomputer that led the Top500 ranking from June 2020 to November 2021. After porting BWA-MEM2 we define and implement a number of optimizations to improve performance in the A64FX target architecture. We show that while the A64FX performance is lower than that of the Skylake system, A64FX delivers 11.6% better energy-to-solution on average. All the code used for this article is available at https://gitlab.bsc.es/rlangari/bwa-a64fx.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Wavefront parallelization of recurrent neural networks on multi-core architectures

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

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

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

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

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

Mont-Blanc 2020: Towards scalable and power efficient European HPC processors

  • Armejach Sanosa, Adrià
  • Brank, Bine
  • Cortina Guardia, Jordi
  • Dolique, François
  • Hayes, Timothy
  • Ho, Nam
  • Lagadec, Pierre-Axel
  • Lemaire, Romain
  • López Paradís, Guillem
  • Marliac, Laurent
  • Moretó Planas, Miquel
  • Marcuello Pascual, Pedro
  • Pleiter, Dirk
  • Tan, Xubin
  • Derradji, Said
The Mont-Blanc 2020 (MB2020) project has triggered the development of the next generation industrial processor for Big Data and High Performance Computing (HPC). MB2020 is paving the way to the future low-power European processor for exascale, defining the System-on-Chip (SoC) architecture and implementing new critical building blocks to be integrated in such an SoC. In this paper, we first present an overview of the MB2020 project, then we describe our experimental infrastructure, the requirements of relevant applications, and the IP blocks developed in the project. Finally, we present our emulation-based final demonstrator and explain how it integrates within our first generation of HPC processors., This work is supported by the European Community’s Horizon 2020 Framework Programme under the Mont-Blanc 2020 project, grant agreement n. 779877., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Characterizing the impact of last-level cache replacement policies on big-data workloads

  • Jamet, Alexandre Valentin
  • Álvarez Martí, Lluc
  • Jiménez, Daniel A.
  • Casas, Marc
The vast disparity between Last Level Cache (LLC) and memory latencies has motivated the need for efficient cache management policies. The computer architecture literature abounds with work on LLC replacement policy. Although these works greatly improve over the least-recently-used (LRU) policy, they tend to focus only on the SPEC CPU 2006 benchmark suite - and more recently on the SPEC CPU 2017 benchmark suite - for evaluation. However, these workloads are representative for only a subset of current High-Performance Computing (HPC) workloads. In this paper we evaluate the behavior of a mix of graph processing, scientific and industrial workloads (GAP, XSBench and Qualcomm) along with the well-known SPEC CPU 2006 and SPEC CPU 2017 workloads on state-of-the-art LLC replacement policies such as Multiperspective Reuse Prediction (MPPPB), Glider, Hawkeye, SHiP, DRRIP and SRRIP. Our evaluation reveals that, even though current state-of-the-art LLC replacement policies provide a significant performance improvement over LRU for both SPEC CPU 2006 and SPEC CPU 2017 workloads, those policies are hardly able to capture the access patterns and yield sensible improvement on current HPC and big data workloads due to their highly complex behavior. In addition, this paper introduces two new LLC replacement policies derived from MPPPB. The first proposed replacement policy, Multi-Sampler Multiperspective (MS-MPPPB), uses multiple samplers instead of a single one and dynamically selects the best-behaving sampler to drive reuse distance predictions. The second replacement policy presented in this paper, Multiperspective with Dynamic Features Selector (DS-MPPPB), selects the best behaving features among a set of 64 features to improve the accuracy of the predictions. On a large set of workloads that stress the LLC, MS-MPPPB achieves a geometric mean speed-up of 8.3% over LRU, while DS-MPPPB outperforms LRU by a geometric mean speedup of 8.0%. For big data and HPC workloads, the two proposed techniques present higher performance benefits than state-of-the-art approaches such as MPPPB, Glider and Hawkeye, which yield geometric mean speedups of 7.0%, 5.0% and 4.8% over LRU, respectively., 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). Lluc Alvarez has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under the Juan de la Cierva Formacion fellowship number FJCI-2016-30984. Marc Casas has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramon y Cajal fellowship number RYC-2017-23269. This research was supported by National Science Foundation grant CCF-1912617, Semiconductor Research Corporation project 2936.001, and generous gifts from Intel Labs., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Design trade-offs for emerging HPC processors based on mobile market technology

  • Armejach Sanosa, Adrià
  • Casas, Marc
  • Moretó Planas, Miquel
This is a post-peer-review, pre-copyedit version of an article published in The Journal of Supercomputing. The final authenticated version is available online at: http://dx.doi.org/10.1007/s11227-019-02819-4, High-performance computing (HPC) is at the crossroads of a potential transition toward mobile market processor technology. Unlike in prior transitions, numerous hardware vendors and integrators will have access to state-of-the-art processor designs due to Arm’s licensing business model. This fact gives them greater flexibility to implement custom HPC-specific designs. In this paper, we undertake a study to quantify the different energy-performance trade-offs when architecting a processor based on mobile market technology. Through detailed simulations over a representative set of benchmarks, our results show that: (i) a modest amount of last-level cache per core is sufficient, leading to significant power and area savings; (ii) in-order cores offer favorable trade-offs when compared to out-of-order cores for a wide range of benchmarks; and (iii) heterogeneous configurations help to improve processor performance and energy efficiency., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

On the benefits of tasking with OpenMP

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

Stencil codes on a vector length agnostic architecture

  • Armejach Sanosa, Adrià
  • Caminal Pallarés, Helena
  • Cebrián González, Juan Manuel
  • González-Alberquilla, Rekai
  • Adeniyi-Jones, Chris
  • Valero Cortés, Mateo
  • Casas, Marc
  • Moretó Planas, Miquel
Data-level parallelism is frequently ignored or underutilized. Achieved through vector/SIMD capabilities, it can provide substantial performance improvements on top of widely used techniques such as thread-level parallelism. However, manual vectorization is a tedious and costly process that needs to be repeated for each specific instruction set or register size. In addition, automatic compiler vectorization is susceptible to code complexity, and usually limited due to data and control dependencies. To address some these issues, Arm recently released a new vector ISA, the Scalable Vector Extension (SVE), which is Vector-Length Agnostic (VLA). VLA enables the generation of binary files that run regardless of the physical vector register length. In this paper we leverage the main characteristics of SVE to implement and optimize stencil computations, ubiquitous in scientific computing. We show that SVE enables easy deployment of textbook optimizations like loop unrolling, loop fusion, load trading or data reuse. Our detailed simulations using vector lengths ranging from 128 to 2,048 bits show that these optimizations can lead to performance improvements over straight-forward vectorized code of up to 56.6% for 2,048 bit vectors. In addition, we show that certain optimizations can hurt performance due to a reduction in arithmetic intensity, and provide insight useful for compiler optimizers., This work has been partially supported by the European HiPEAC Network of Excellence, by the Spanish Ministry of Economy and Competitiveness (contract TIN2015-65316-P), and by the Generalitat de Catalunya (contracts 2017-SGR-1328 and 2017-SGR-1414). The Mont-Blanc project receives funding from the EUs H2020 Framework Programme (H2020/2014-2020) under grant agreements no. 671697 and no. 779877. M. Moreto has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramon y Cajal fellowship number RYC-2016-21104. Finally, A. Armejach has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Juan de la Cierva postdoctoral fellowship number FJCI-2015-24753., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Teaching HPC systems and parallel programming with small-scale clusters

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

Semi-automatic validation of cycle-accurate simulation infrastructures: The case for gem5-x86

  • Cebrián González, Juan Manuel
  • Barredo Ferreira, Adrián
  • Caminal Pallarés, Helena
  • Moretó Planas, Miquel
  • Casas, Marc
  • Valero Cortés, Mateo
Since the early 70s, simulation infrastructures have been a keystone in computer architecture research, providing a fast and reliable way to prototype and evaluate ideas for future computing systems. There are different types of simulators, from most detailed (cycle-accurate) to time-based/functional and analytical modeling. Increasing accuracy translates into several orders of magnitude in terms of simulation speed. Yet, a question remains open: are the results derived from the simulation infrastructure representative of a real machine? Validation of these infrastructures is complex and costly, usually performed upon release. However, most simulators do not provide the appropriate means to verify or validate new architectural models. In this paper, we introduce a semi-automatic validation framework based on real-hardware performance counter information. The framework provides two levels of abstraction: (a) a high level definition of the processor behavior (Top-Down model) and (b) detailed per-structure and per-pipeline-stage usage breakdown to pinpoint simulator issues. We used this framework to validate the latest available gem5-x86 simulation environment, and found several sources of error that alter the expected behavior of the simulated processor, which we were later to document and correct., This work has been partially supported by the Spanish Government (Severo Ochoa grants SEV2015-0493, SEV-2011-00067), the Spanish Ministry of Science and Innovation (contract TIN2015- 65316-P), Generalitat de Catalunya, Spain (contracts 2014-SGR1051 and 2014-SGR-1272), the RoMoL ERC Advanced, Spain Grant (GA 321253), the European HiPEAC Network of Excellence and the Mont-Blanc project (EU-FP7-610402 and EU-H2020-779877). A. Barredo has been supported by the Spanish Government under Formación del Personal Investigador fellowship number BES2017-080635. M. Moreto and M. Casas have been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness, Spain under Ramon y Cajal fellowship numbers RYC2016-21104 and RYC-2017-23269., Peer Reviewed, Postprint (author's final draft)
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 Contribución de congreso (ConferenceOutput)

BST: A BookSim-based toolset to simulate NoCs with single- and multi-hop bypass

  • Pérez Gallardo, Iván
  • Vallejo Gutiérrez, Enrique
  • Moretó Planas, Miquel
  • Beivide Palacio, Julio Ramon
Network-on-Chips are a critical part of modern multiprocessors and their relevance will grow with the number of cores. The development of future NoC designs relies on detailed simulation models that accurately estimate their performance, power and hardware cost. Bypass routers are promising proposals due to their improved performance. Bypass routers reduce latency thanks to a combination of speculation, pre-routing (lookahead routing) and buffer bypass, which also reduce energy consumption by avoiding unnecessary buffer writes and reads. Multi-hop bypass NoCs, known as SMART, even bypass the crossbar of multiple routers in a single cycle. However, publicly available NoC simulators, such as BookSim or Garnet, do not implement bypass mechanisms or do not model them accurately. In this work, we present Bypass Simulation Toolset (BST), a set of tools to accurately simulate NoCs with single- and multi-hop bypass routers. BST combines and extends several simulation tools: an extension of BookSim with state-of-theart cycle-accurate bypass router models and additional flow control mechanisms; an RTL implementation of multi-hop bypass mechanisms based on OpenSMART; an API to ease a modular integration of the BST NoC simulator in full system simulators; and a set of scripts to automate simulation execution and data collection. To showcase BST, we i) validate BookSim SMART models with the RTL implementation; ii) compare bypass and traditional nonbypass router models; iii) integrate BookSim in gem5 using the proposed API and compare it with gem5’s Simple and Garnet 2.0 NoC models; and iv) present a case study evaluating different combinations of router types and topologies recently proposed for NoCs, highlighting the flexibility of the BST toolset., This work was supported by the Spanish Ministry of Science, Innovation and Universities, contracts TIN2015-65316- P, TIN2016-76635-C2-2-R (AEI/FEDER, UE) and PID2019- 105660RB-C22; the European Union’s Horizon 2020 research and innovation program under the Mont-Blanc 2020 project (grant agreement 779877); and the HiPEAC Network of Excellence. I. Perez is partially supported by an FPI grant, BES- ´ 2017-079971. M. Moreto has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramon y Cajal fellowship number RYC-2016-21104. Bluespec Inc. provided access to Bluespec tools., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación

Towards resilient EU HPC systems: A blueprint

  • Radojković, Petar
  • Marazakis, Manolis
  • Carpenter, Paul Matthew
  • Jeyapaul, Reiley
  • Gizopoulos, Dimitris
  • Schulz, Martin
  • Armejach Sanosa, Adrià
  • Ayguadé Parra, Eduard
  • Canal Corretger, Ramon
  • Moretó Planas, Miquel
  • Salami, Behzad
  • Unsal, Osman Sabri
This document aims to spearhead a Europe-wide discussion on HPC system resilience and to help the European HPC community define best practices for resilience. We analyse a wide range of state-of-the-art resilience mechanisms and recommend the most effective approaches to employ in large-scale HPC systems. Our guidelines will be useful in the allocation of available resources, as well as guiding researchers and research funding towards the enhancement of resilience approaches with the highest priority and utility. Although our work is focused on the needs of next generation HPC systems in Europe, the principles and evaluations are applicable globally., This work has received funding from the European Union’s Horizon 2020 research and innovation programme under the projects ECOSCALE (grant agreement No 671632), EPI (grant agreement No 826647), EuroEXA (grant agreement No 754337), Eurolab4HPC (grant agreement No 800962), EVOLVE (grant agreement No 825061), EXA2PRO (grant agreement No 801015), ExaNest (grant agreement No 671553), ExaNoDe (grant agreement No 671578), EXDCI-2 (grant agreement No 800957), LEGaTO (grant agreement No 780681), MB2020 (grant agreement No 779877), RECIPE (grant agreement No 801137) and SDK4ED (grant agreement No 780572). The work was also supported by the European Commission’s Seventh Framework Programme under the projects CLERECO (grant agreement No 611404), the NCSA-Inria-ANL-BSC-JSCRiken-UTK Joint-Laboratory for Extreme Scale Computing – JLESC (https://jlesc.github.io/), OMPI-X project (No ECP-2.3.1.17) and the Spanish Government through Severo Ochoa programme (SEV-2015-0493). This work was sponsored in part by the U.S. Department of Energy's Office of Advanced Scientific Computing Research, program managers Robinson Pino and Lucy Nowell. This manuscript has been authored by UT-Battelle, LLC under Contract No DE-AC05-00OR22725 with the U.S. Department of Energy., Preprint
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Data prefetching on in-order processors

  • Ortega Carrasco, Cristobal
  • García Flores, Víctor
  • Moretó Planas, Miquel
  • Casas, Marc
  • Rositoru, Roxana
Low-power processors have attracted attention due to their energy-efficiency. A large market, such as the mobile one, relies on these processors for this very reason. Even High Performance Computing (HPC) systems are starting to consider low-power processors as a way to achieve exascale performance within 20MW, however, they must meet the right performance/Watt balance. Current low-power processors contain in-order cores, which cannot re-order instructions to avoid data dependency-induced stalls. Whilst this is useful to reduce the chip's total power consumption, it brings several challenges. Due to the evolving performance gap between memory and processor, memory is a significant bottleneck. In-order cores cannot re-order instructions and are memory latency bound, something data prefetching can help alleviate by ensuring data is readily available. In this work, we do an exhaustive analysis of available data prefetching techniques in state-of-The-Art in-order cores. We analyze 5 static prefetchers and 2 dynamic aggressiveness and destination mechanisms applied to 3 data prefetchers on a set of HPC mini-and proxy-Applications, whilst running on in-order processors. We show that next-line prefetching can achieve nearly top performance with a reasonable bandwidth consumption when throttled, whilst neighbor prefetchers have been found to be best, overall., This work has been supported by the RoMoL ERC Advanced Grant (GA 321253), by the European HiPEAC Network of Excellence, by the Spanish Ministry of Science and Innovation (contracts TIN2015-65316-P), by Generalitat de Catalunya (contracts 2014-SGR-1051 and 2014-SGR-1272), by the European Union’s Horizon 2020 research and innovation programme (grant agreements No 671697 and No 779877). M. Moreto has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramon y Cajal fellowship number RYC-2016-21104., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

PrioRAT: criticality-driven prioritization inside the on-chip memory hierarchy

  • Dimic, Vladimir
  • Moretó Planas, Miquel
  • Casas, Marc
  • Valero Cortés, Mateo
The ever-increasing gap between the processor and main memory speeds requires careful utilization of the limited memory link. This is additionally emphasized for the case of memory-bound applications. Prioritization of memory requests in the memory controller is one of the approaches to improve performance of such codes. However, current designs do not consider high-level information about parallel applications. In this paper, we propose a holistic approach to this problem, where the runtime system-level knowledge is made available in hardware. Processor exploits this information to better prioritize memory requests, while introducing negligible hardware cost. Our design is based on the notion of critical path in the execution of a parallel code. The critical tasks are accelerated by prioritizing their memory requests within the on-chip memory hierarchy. As a result, we reduce the critical path and improve the overall performance up to 1.19× compared to the baseline systems., This work has been partially supported by the Spanish Ministry of Science and Innovation (PID2019-107255GB-C21/AEI/10.13039/ 501100011033), by the Generalitat de Catalunya (contracts 2017-SGR-1414 and 2017-SGR-1328), by the European Unions Horizon 2020 research and innovation program under the Mont-Blanc 2020 project (grant agreement 779877) and by the RoMoL ERC Advanced Grant (GA 321253). V. Dimić has been partially supported by the Agency for Management of University and Research Grants (AGAUR) of the Government of Catalonia under Ajuts per a la contractaci o de personal investigador novell fellowship number 2017 FI B 00855. M. Moreto and M. Casas have been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramon y Cajal fellowship numbers RYC- 2016-21104 and RYC-2017-23269, respectively., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

TaskGenX: A Hardware-Software Proposal for Accelerating Task Parallelism

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

Demystifying the Characteristics of High Bandwidth Memory for Real-Time Systems

  • Asifuzzaman, Kazi
  • Abuelala, Mohamed
  • Hassan, Mohamed
  • Cazorla Almeida, Francisco Javier
The number of functionalities controlled by software on every critical real-time product is on the rise in domains like automotive, avionics and space. To implement these advanced functionalities, software applications increasingly adopt artificial intelligence algorithms that manage massive amounts of data transmitted from various sensors. This translates into unprecedented memory performance requirements in critical systems that the commonly used DRAM memories struggle to provide. High-Bandwidth Memory (HBM) can satisfy these requirements offering high bandwidth, low power and high-integration capacity features. However, it remains unclear whether the predictability and isolation properties of HBM are compatible with the requirements of critical embedded systems. In this work, we perform to our knowledge the first timing analysis of HBM. We show the unique structural and timing characteristics of HBM with respect to DRAM memories and how they can be exploited for better time predictability, with emphasis on increased isolation among tasks and reduced worst-case memory latency., This work has been partially supported by the Spanish Ministry of Science and Innovation under grant PID2019-107255GB-C21/AEI/10.13039/501100011033; the European Union’s Horizon 2020 Framework Programme under grant agreement No. 878752 (MASTECS) and agreement No. 779877 (Mont-Blanc 2020); the European Research Council (ERC) grant agreement No. 772773 (SuPerCom); and the Natural Sciences and Engineering Research Council of Canada (NSERC), Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

A vulnerability factor for ECC-protected memory

  • Jaulmes, Luc
  • Moretó Planas, Miquel
  • Valero Cortés, Mateo
  • Casas, Marc
Fault injection studies and vulnerability analyses have been used to estimate the reliability of data structures in memory. We survey these metrics and look at their adequacy to describe the data stored in ECC-protected memory. We also introduce FEA, a new metric improving on the memory derating factor by ignoring a class of false errors. We measure all metrics using simulations and compare them to the outcomes of injecting errors in real runs. This in-depth study reveals that FEA provides more accurate results than any state-of-the-art vulnerability metric. Furthermore, FEA gives an upper bound on the failure probability due to an error in memory, making this metric a tool of choice to quantify memory vulnerability. Finally, we show that ignoring these false errors reduces the failure rate on average by 12.75% and up to over 45%., This work has been supported by the RoMoL ERC Advanced Grant (GA 321253), by the European HiPEAC Network of Excellence, by the Spanish Ministry of Economy and Competitiveness (contract TIN2015-65316- P), by the Generalitat de Catalunya (contracts 2017-SGR-1414 and 2017- SGR-1328), by the Spanish Government (Severo Ochoa grant SEV-2015- 0493) and by the European Union’s Horizon 2020 research and innovation programme (grant agreements 671697 and 779877). L. Jaulmes has been partially supported by the Spanish Ministry of Education, Culture and Sports under grant FPU2013/06982. M. Moreto and M. Casas have been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramon y Cajal fellowships RYC-2016-21104 and RYC-2017-23269., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

On the maturity of parallel applications for asymmetric multi-core processors

  • Chronaki, Kallia
  • Moretó Planas, Miquel
  • Casas, Marc
  • Rico, Alejandro
  • Badia Sala, Rosa Maria
  • Ayguadé Parra, Eduard
  • Valero Cortés, Mateo
Asymmetric multi-cores (AMCs) are a successful architectural solution for both mobile devices and supercomputers. By maintaining two types of cores (fast and slow) AMCs are able to provide high performance under the facility power budget. This paper performs the first extensive evaluation of how portable are the current HPC applications for such supercomputing systems. Specifically we evaluate several execution models on an ARM big.LITTLE AMC using the PARSEC benchmark suite that includes representative highly parallel applications. We compare schedulers at the user, OS and runtime levels, using both static and dynamic options and multiple configurations, and assess the impact of these options on the well-known problem of balancing the load across AMCs. Our results demonstrate that scheduling is more effective when it takes place in the runtime system level as it improves the baseline by 23%, while the heterogeneous-aware OS scheduling solution improves the baseline by 10%., This work has been supported by the RoMoL ERC Advanced Grant (GA 321253), by the European HiPEAC Network of Excellence, by the Spanish Ministry of Science and Innovation (contracts TIN2015-65316-P), by the Generalitat de Catalunya (contracts 2014-SGR-1051 and 2014-SGR-1272), and by the European Union's Horizon 2020 research and innovation programme under grant agreement No 671697 and No. 779877. M. Moretó has been partially supported by the Ministry of Economy and Competitiveness under Ramon y Cajal fellowship number RYC-2016-21104., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Design space exploration of next-generation HPC machines

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

gem5 + rtl: A framework to enable RTL models inside a full-system simulator

  • López Paradís, Guillem
  • Armejach Sanosa, Adrià
  • Moretó Planas, Miquel
In recent years there has been a surge of interest in designing custom accelerators for power-efficient high-performance computing. However, available tools to simulate low-level RTL designs often neglect the target system in which the design will operate. This hinders proper testing and debugging of functionalities, and does not allow co-designing the accelerator to obtain a balanced and efficient architecture. In this paper, we introduce gem5 + rtl, a flexible framework that enables simulation of RTL models inside a full-system software simulator. We present the framework’s functionality that allows easy integration of RTL models on a simulated system-on-chip (SoC) that is able to boot Linux and run complex multi-threaded and multi-programmed workloads. We demonstrate the framework with two relevant use cases that integrate a multi-core SoC with a Performance Monitoring Unit (PMU) and the NVIDIA Deep Learning Accelerator (NVDLA), showcasing how the framework enables testing RTL model features and how it can enable co-design taking into account the entire SoC., This research was supported by the European Union Regional Development Fund within the framework of the ERDF Operational Program of Catalonia 2014-2020 with a grant of 50% of total cost eligible under the DRAC project [001-P- 001723], by the Spanish goverment (grant RTI2018-095094- B-C21 CONSENT), by the Spanish Ministry of Science and Innovation (contracts PID2019-107255GB-C21) and by the Catalan Government (contracts 2017-SGR-1414, 2017-SGR705). This work has also been supported by the European Community’s Horizon 2020 Framework Programme under the Mont-Blanc 2020 and EPI projects (grant agreements n. 779877 and n. 826647); and by the Arm-BSC Center of Excellence. G. López-Paradís has been partially supported by the Agency for Management of University and Research Grants (AGAUR) of the Government of Catalonia under Ajuts per a la contractació de personal investigador novell fellowship No. 2021FI B00994. A. Armejach has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Juan de la Cierva postdoctoral fellowship number IJCI-2017-33945. M. Moretó has been partially supported by the Spanish Ministry of Economy, Industry and Competitiveness under Ramón y Cajal fellowship No. RYC-2016-21104., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Improving predication efficiency through compaction/restoration of SIMD instructions

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

Runtime-assisted cache coherence deactivation in task parallel programs

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