Pasar al contenido principal

HETEROGENEIDAD Y ESPECIALIZACION EN LA ERA POST-MOORE

RTI2018-093684-B-I00

•
Nombre agencia financiadora Agencia Estatal de Investigación
Acrónimo agencia financiadora AEI
Programa Programa Estatal de I+D+i Orientado a los Retos de la Sociedad
Convocatoria Proyectos I+D+i Retos de Investigación
Año convocatoria 2018
Plan Estatal 2017-2020 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario UNIVERSIDAD COMPLUTENSE DE MADRID

Resultados relacionados

Resultados totales (Incluyendo duplicados): 13
Encontrada(s) 1 página(s)

Docta Complutense
Publicación Contribución de congreso (ConferenceOutput)

Enabling performance portability of data-parallel OpenMP applications on asymmetric multicore processors

  • Sáez Alcaide, Juan Carlos
  • Castro Rodríguez, Fernando
  • Prieto Matías, Manuel
Asymmetric multicore processors (AMPs) couple high-performance big cores and low-power small cores with the same instruction-set architecture but different features, such as clock frequency or microarchitecture. Previous work has shown that asymmetric designs may deliver higher energy efficiency than symmetric multicores for diverse workloads. Despite their benefits, AMPs pose significant challenges to runtime systems of parallel programming models. While previous work has mainly explored how to efficiently execute task-based parallel applications on AMPs, via enhancements in the runtime system, improving the performance of unmodified data-parallel applications on these architectures is still a big challenge. In this work we analyze the particular case of loop-based OpenMP applications, which are widely used today in scientific and engineering domains, and constitute the dominant application type in many parallel benchmark suites used for performance evaluation on multicore systems. We observed that conventional loop-scheduling OpenMP approaches are unable to efficiently cope with the load imbalance that naturally stems from the different performance delivered by big and small cores. To address this shortcoming, we propose Asymmetric Iteration Distribution (AID), a set of novel loop-scheduling methods for AMPs that distribute iterations unevenly across worker threads to efficiently deal with performance asymmetry. We implemented AID in libgomp –the GNU OpenMP runtime system–, and evaluated it on two different asymmetric multicore platforms. Our analysis reveals that the AID methods constitute effective replacements of the static and dynamic methods on AMPs, and are capable of improving performance over these conventional strategies by up to 56% and 16.8%, respectively.
Docta Complutense
Publicación Artículo científico (JournalArticle)

LFOC+: A Fair OS-Level Cache-Clustering Policy for Commodity Multicore Systems

  • Sáez Alcaide, Juan Carlos
  • Castro Rodríguez, Fernando
  • Fanizzi, Graziano
  • Prieto Matías, Manuel
Commodity multicore systems are increasingly adopting hardware support that enables the system software to partition the last-level cache (LLC). This support makes it possible for the operating system (OS) or the Virtual Machine Monitor (VMM) to mitigate shared-resource contention effects on multicores by assigning different co-running applications to various cache partitions. Recently cache-clustering (or partition-sharing) strategies have emerged as a way to improve system throughput and fairness on new platforms with cache-partitioning support. As opposed to strict cache-partitioning, which allocates separate cache partitions to each application, cache-clustering allows partitions to be shared by a group of applications. In this article we propose LFOC+, a fairness-aware OS-level cache-clustering policy for commodity multicore systems. LFOC+ tries to mimic the behavior of the optimal cache-clustering solution for fairness, which we could obtain for different workload scenarios by using a simulation tool. Our dynamic cache-clustering strategy continuously gathers data fromperformancemonitoring counters to classify applications at runtime based on the degree of cache sensitivity and contentiousness, and effectively separates cache-sensitive applications fromaggressor programs to improve fairness,while providing acceptable system throughput.We implemented LFOC+ in the Linux kernel and evaluated it on a real systemfeaturing an Intel Skylake processor, wherewe compare its effectiveness to that of four previously proposed cache-clustering policies. Our experimental análisis reveals that LFOC+ constitutes a lightweight OS-level policy and improves fairness relative to two other state-of-the-art fairness-aware strategies –Dunn and LFOC–, by up to 22% and up to 20.6%, respectively, and by9% and 4.9%on average.
Docta Complutense
Publicación Contribución de congreso (ConferenceOutput)

LFOC: A Lightweight Fairness-Oriented Cache Clustering Policy for Commodity Multicores

  • García García, Adrián
  • Sáez Alcaide, Juan Carlos
  • Castro Rodríguez, Fernando
  • Prieto Matías, Manuel
Multicore processors constitute the main architecture choice for modern computing systems in different market segments. Despite their benefits, the contention that naturally appears when multiple applications compete for the use of shared resources among cores, such as the last-level cache (LLC), may lead to substantial performance degradation. This may have a negative impact on key system aspects such as throughput and fairness. Assigning the various applications in the workload to separate LLC partitions with possibly different sizes, has been proven effective to mitigate shared-resource contention effects. In this article we propose LFOC, a clustering-based cache partitioning scheme that strives to deliver fairness while providing acceptable system throughput. LFOC leverages the Intel Cache Allocation Technology (CAT), which enables the system software to divide the LLC into different partitions. To accomplish its goals, LFOC tries to mimic the behavior of the optimal cache-clustering solution, which we could approximate by means of a simulator in different scenarios. To this end, LFOC effectively identifies streaming aggressor programs and cache sensitive applications, which are then assigned to separate cache partitions. We implemented LFOC in the Linux kernel and evaluated it on a real system featuring an Intel Skylake processor, where we compare its effectiveness to that of two state-of-the-art policies that optimize fairness and throughput, respectively. Our experimental analysis reveals that LFOC is able to bring a higher reduction in unfairness by leveraging a lightweight algorithm suitable for adoption in a real OS.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Anatomy of the BLIS family of algorithms for matrix multiplication

  • Castelló, Adrián|||0000-0002-8576-8451
  • Quintana-Ortí, Enrique S.|||0000-0002-5454-165X
  • Igual, Francisco D.
[EN] The efforts of the scientific community and hardware vendors to develop and optimize linear algebra codes have historically led to highly-tuned libraries, carefully adapted to the underlying processor architecture, with excellent (near-peak) performance. These optimization efforts, however, are commonly focused on obtaining the best performance possible when the involved operands are large and "squarish" matrices. New computationally-intensive applications (e.g., in deep learning) are increasingly demanding high-performance BLAS (Basic Linear Algebra Subprograms) also for small operands in any of their dimensions. In this paper, we tackle this problem by refactoring the general matrix-matrix multiplication (GEMM) algorithm within a specific high-performance implementation of BLAS, named BLIS, proposing a complete family of algorithmic variants to implement GEMM with different strategies to exploit the target cache hierarchy, together with the changes to be applied to architecture-specific codes to instantiate a complete GEMM implementation. Experimental results on an ARM processor (NVIDIA Carmel) reveal significant performance differences between the members of the GEMM family, depending on the shape and dimension of the matrix operands., This work was supported by the research project PID2020-113656RB-C22 of MCIN/AEI/10.13039/501100011033, EU (FEDER), the Spanish MINECO, under grants RTI2018-093684-B-I00, PID2021-126576NB-I00, and CM via Multiannual Agreement with Complutense University in the line Program to Stimulate Research for Young Doctors in the context of the V PRICIT under projects PR65/19-22445 and CM S2018/TCS-4423. A. Castello is a FJC2019-039222-I fellow supported by MCIN/AEI/10.13039/501100011033.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Micro-kernels for portable and efficient matrix multiplication in deep learning

  • Alaejos-López, Guillermo
  • Martínez, Héctor
  • Igual, Francisco D.
  • Castelló, Adrián|||0000-0002-8576-8451
  • Alonso-Jordá, Pedro|||0000-0002-6882-6592
  • Quintana-Ortí, Enrique S.|||0000-0002-5454-165X
[EN] We provide a practical demonstration that it is possible to systematically generate a variety of high-performance micro-kernels for the general matrix multiplication (gemm) via generic templates which can be easily customized to different processor architectures and micro-kernel dimensions. These generic templates employ vector intrinsics to exploit the SIMD (single instruction, multiple data) units in current general-purpose processors and, for the particular type of gemm problems encountered in deep learning, deliver a floating-point throughput rate on par with or even higher than that obtained with conventional, carefully tuned implementations of gemm in current linear algebra libraries (e.g., BLIS, AMD AOCL, ARMPL). Our work exposes the structure of the template-based micro-kernels for ARM Neon (128-bit SIMD), ARM SVE (variable-length SIMD) and Intel AVX512 (512-bit SIMD), showing considerable performance for an NVIDIA Carmel processor (ARM Neon), a Fujitsu A64FX processor (ARM SVE) and on an AMD EPYC 7282 processor (256-bit SIMD)., Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature. This work was supported by the research projects PID2020-113656RB-C22, RTI2018-093684-B-I00, and PID2021-126576NB-I00, of MCIN/AEI/10.13039/501100011033 and by "ERDF A way of making Europe", and CM via Multiannual Agreement with Complutense University in the line Program to Stimulate Research for Young Doctors in the context of the V PRICIT under projects PR65/19-22445 and CM S2018/TCS-4423. A. Castello is a FJC2019-039222-I fellow supported by MCIN/AEI/10.13039/501100011033. H. Martinez is a postdoctoral fellow supported by the Consejeria de Transformacion Economica, Industria, Conocimiento y Universidades de la Junta de Andalucia.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Low precision matrix multiplication for efficient deep learning in NVIDIA Carmel processors

  • San Juan-Sebastian, Pablo
  • Rodríguez-Sánchez, Rafael
  • Igual, Francisco D.
  • Alonso-Jordá, Pedro|||0000-0002-6882-6592
  • Quintana-Ortí, Enrique S.|||0000-0002-5454-165X
[EN] We introduce a high performance, multi-threaded realization of the gemm kernel for the ARMv8.2 architecture that operates with 16-bit (half precision)/queryKindly check and confirm whether the corresponding author is correctly identified. floating point operands. Our code is especially designed for efficient machine learning inference (and to a certain extent, also training) with deep neural networks. The results on the NVIDIA Carmel multicore processor, which implements the ARMv8.2 architecture, show considerable performance gains for the gemm kernel, close to the theoretical peak acceleration that could be expected when moving from 32-bit arithmetic/data to 16-bit. Combined with the type of convolution operator arising in convolutional neural networks, the speed-ups are more modest though still relevant., This work was supported by projects TIN2017-82972-R and RTI2018-093684-B-I00 from the Ministerio de Ciencia, Innovacion y Universidades, project S2018/TCS-4423 of the Comunidad de Madrid, project PR65/19-22445 of the UCM, and project Prometeo/2019/109 of the Generalitat Valenciana.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Integration and exploitation of intra-routine malleability in BLIS

  • Rodríguez-Sánchez, Rafael
  • Igual, Francisco D.
  • Quintana-Ortí, Enrique S.|||0000-0002-5454-165X
[EN] Malleability is a property of certain applications (or tasks) that, given an external request or autonomously, can accommodate a dynamic modification of the degree of parallelism being exploited at runtime. Malleability improves resource usage (core occupation) on modern multicore architectures for applications that exhibit irregular and divergent execution paths and heavily depend on the underlying library performance to attain high performance. The integration of malleability within high-performance instances of the Basic Linear Algebra Subprograms (BLAS) is nonexistent, and, in addition, it is difficult to attain given the rigidity of current application programming interfaces (APIs). In this paper, we overcome these issues presenting the integration of a malleability mechanism within BLIS, a high-performance and portable framework to implement BLAS-like operations. For this purpose, we leverage low-level (yet simple) APIs to integrate on-demand malleability across all Level-3 BLAS routines, and we demonstrate the performance benefits of this approach by means of a higher-level dense matrix operation: the LU factorization with partial pivoting and look-ahead, The researchers from Universidad Complutense de Madrid were supported by the EU (FEDER) and Spanish MINECO (TIN2015-65277-R, RTI2018-093684-B-I00), and by Spanish CM (S2018/TCS-4423). The researcher from Universitat Poliecnica de Valencia was supported by the Spanish MINECO (TIN2017-82972-R)
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Fine-grain task-parallel algorithms for matrix factorizations and inversion on many-threaded CPUs

  • Catalán, Sandra
  • Herrero, José R.
  • Igual, Francisco D.
  • Rodríguez-Sánchez, Rafael
  • Quintana-Ortí, Enrique S.|||0000-0002-5454-165X
[EN] We extend a two-level task partitioning previously applied to the inversion of dense matrices via Gauss-Jordan elimination to the more challenging QR factorization as well as the initial orthogonal reduction to band form found in the singular value decomposition. Our new task-parallel algorithms leverage the tasking mechanism currently available in OpenMP to exploit "nested" task parallelism, with a first outer level that operates on matrix panels and a second inner level that processes the matrix either by mu$$ \mu $$-panels or by tiles, in order to expose a large number of independent tasks. We present a detailed performance analysis, including execution traces, which shows that the two-level refinement into fine grain tasks allows for an improved load balancing and delivers high performance on current general-purpose many-core processors (CPUs) from Intel and AMD., This research was sponsored by projects RTI2018-093684-B-I00, PID2019-107255GB and TIN2017-82972-R ofMinisterio de Ciencia, Innovación y Universidades; project S2018/TCS-4423 of Comunidad de Madrid; project 2017-SGR-1414 of the Generalitat de Catalunya and the Madrid Government under the Multiannual Agreement with UCM in the line Program to Stimulate Research for Young Doctors in the context of the V PRICIT, project PR65/19-22445.
Docta Complutense
Publicación Contribución de congreso (ConferenceOutput)

Rapid Development of OS Support with PMCSched for Scheduling on Asymmetric Multicore Systems

  • Bilbao, Carlos
  • Sáez Alcaide, Juan Carlos
  • Prieto Matías, Manuel
Asymmetric multicore processors (AMPs) couple high-performance big cores and power-efficient small ones, all exposing a shared instruction set architecture to software, but with different microarchitectural features. The energy efficiency benefits of AMPs together with the general-purpose nature of the various cores, have led hardware manufactures to build commercial AMP-based products, first for the mobile and embedded domains, and more recently for the desktop market segment, as with the Intel Alder Lake processor family. This indicates that AMPs may become a solid and more energy efficient replacement to symmetric multicores in a wide range of application domains. Previous research has demonstrated that the system software can substantially improve scheduling –critical to get the most out of heterogeneous cores– by leveraging hardware facilities that are directly managed by the OS, such as performance monitoring counters, or the recently introduced Intel Thread Director technology. Unfortunately, the OS-level support enabling to access scheduling-relevant hardware support may take a long time to be adopted in operating systems, or may come in forms that make its utilization challenging from specific levels of the system software stack, especially in production systems. To fill this gap, we propose the PMCSched framework, which enables the creation of custom OS support on Linux to aid in the design of novel scheduling and resource-management policies for multicores implemented at different layers of the system software, but without requiring to patch the kernel. To demonstrate the potential of our framework, we implement a set of OS-level schedulers for AMPs, that make use of custom OS extensions to access scheduling-relevant hardware facilities in an x86 AMP processor.
Docta Complutense
Publicación Artículo científico (JournalArticle)

Optimization of a Line detection algorithm for autonomous vehicles on a RISC-V with accelerator

  • Belda Beneyto, María José
  • Olcoz Herrero, Katzalin
  • Castro Rodríguez, Fernando
  • Tirado Fernández, José Francisco
In recent years, autonomous vehicles have attracted the attention of many research groups, both in academia and business, including researchers from leading companies such as Google, Uber and Tesla. This type of vehicles are equipped with systems that are subject to very strict requirements, essentially aimed at performing safe operations -both for potential passengers and pedestrians- as well as carrying out the processing needed for decision making in real time. In many instances, general-purpose processors alone cannot ensure that these safety, reliability and real-time requirements are met, so it is common to implement paper explores the acceleration of a line detection aprunning without accelerator.
Docta Complutense
Publicación Artículo científico (JournalArticle)

PLAM: a posit logarithm-approximate multiplier

  • Murillo Montero, Raúl
  • Del Barrio García, Alberto Antonio
  • Botella Juan, Guillermo
  • Kim, Min Soo
  • Kim, HyunJin
  • Bagherzadeh, Nader
The Posit™ Number System was introduced in 2017 as a replacement for floating-point numbers. Since then, the community has explored its application in several areas, such as deep learning, and produced some unit designs which are still far from being competitive with their floating-point counterparts. This article proposes a Posit Logarithm-Approximate Multiplication (PLAM) scheme to significantly reduce the complexity of posit multipliers, one of the most power-hungry arithmetic units. The impact of this approach is evaluated in deep neural network inference, where there are no significant accuracy drops. Compared with state-of-the-art posit multipliers, experiments show that the proposed technique reduces the area, power, and delay of 32-bit hardware multipliers up to 72.86%, 81.79%, and 17.01%, respectively.
Docta Complutense
Publicación Contribución de congreso (ConferenceOutput)

Evaluation of the Intel Thread Director Technology on an Alder Lake processor

  • Sáez Alcaide, Juan Carlos
  • Prieto Matías, Manuel
Asymmetric multicore processors (AMPs) combine high-performance big cores with more energy-efficient small cores, all exposing a shared instruction-set architecture but different features, such as clock frequency or microarchitecture. In the last decade, most commercial AMP products have mainly targetted the embedded and mobile domains. Today, major hardware players are releasing new AMP-based products that aim to move beyond the mobile niche, towards the desktop/server segments. The Apple M1 SoC or the recent Intel Alder Lake processor family are clear examples of these new AMP systems. Despite their energy-efficiency benefits, AMPs pose significant challenges to the operating system scheduler. In this paper, we assess the effectiveness of the Thread Director (TD) technology, a set of hardware facillities - first introduced in Alder Lake processors - that provide the OS with hints on the performance and energy efficiency that a thread delivers when running on the various core types. The main focus of our analysis is to evaluate how effectively the OS can drive scheduling decisions with TD's performance hints. To this end, we incorporated support in Linux to conveniently access TD facillites from the OS kernel. Motivated by various TD's limitations identified with our analysis, we opted to build hardware-counter based prediction models (generated via machine-learning methods) to better aid the OS in making throughput-oriented and fairness-aware scheduling decisions. The effectiveness of both TD and the hardware-counter based models for performance prediction is evaluated both via offline monitoring, and also online, by utilizing our implementation of various asymmetry-aware schedulers in the Linux kernel.
Docta Complutense
Publicación Contribución de congreso (ConferenceOutput)

Decomposition of Dillon’s APN permutation with efficient hardware implementation

  • Imaña Pascual, José Luis
  • Budaghyan, Lilya
  • Kaleyski, Nikolay
Modern block ciphers incorporate a vectorial Boolean function (S-box ) as their only nonlinear component. Almost Perfect Nonlinear (APN) functions exhibit optimal resistance to differential cryptanalysis and thus present ideal security properties as S-boxes. These optimal cryptographic properties have the side effect of making the function harder to represent and implement. As the number of variables of the function grows, lookup-table representations become less feasible, and so from a practical point of view, it is crucial to develop a good understanding of how cryptographically strong functions can be represented in hardware. This paper focuses on one of the most important APN functions, namely Dillon’s permutation in dimension 6. This is the only known APN permutation in an even number of variables. It is thus an ideal candidate for studying the efficiency of different representations since it combines at least two very important cryptographic properties, and since the number of variables is not large enough to make its computational investigation intractable. In this paper, we give a new description of Dillon’s permutation as a composition of two functions and compare it with its classic univariate polynomial representation. We give hardware architectures for both representations, and we report on the results obtained from their FPGA implementations. From the experimental results, the implementation of the new decomposed Dillon’s permutation presents reductions in the number of 2-input XOR gates of up to 27.3% and in the Area × Delay metrics of up to 27.4% with respect to the implementation of the corresponding univariate representation. Therefore, the new decomposed Dillon’s permutation representation is more efficient than the univariate polynomial one when reconfigurable devices are used for the hardware implementation. This indicates that by representing APN functions as a composition of simpler functions, significant reductions in the complexity of the implementation can be achieved.