Skip to main content

Resultados relacionados

Resultados totales (Incluyendo duplicados): 6
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

SMT Efficiency in Supervised ML Methods: a Throughput and Interference Analysis

  • Pons-Escat, Lucía|||0000-0002-4582-7744
  • Navarro-Edo, Marta|||0000-0002-4334-460X
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Pons Terol, Julio|||0000-0002-5654-6753
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
[EN] The microarchitecture of general-purpose processors is continuously evolving to adapt to the new computation and memory demands of incoming workloads. In this regard, new circuitry is added to execute specific instructions like vector multiplication or string operations. These enhancements and the support of multiple threads per core make simultaneous multithreading (SMT) processors dominate the market for data center processors. Regarding emerging workloads, machine learning is taking an important role in many research domains like biomedicine, economics, and social sciences. This paper analyzes the efficiency of machine learning workloads running in SMT mode (two threads per core) versus running them in ST mode (single-threaded) with twice the number of cores. Experimental results in an Intel Xeon Skylake-X processor show an SMT efficiency falling between 80% and 100% across the studied workloads. These results prove two main findings: i) last-generation SMT processors are excellent candidates to execute ML workloads as they achieve a high SMT efficiency, and ii) if the performance of two major resources (i.e., FP double operator and core's caches) was boosted, all the workloads would achieve an almost perfect SMT efficiency. Moreover, results show that there is still room to support more threads without adding extra hardware. The discussed findings are aimed at providing insights to design future processors for ML workloads., This work has been supported by the Spanish Ministerio de Ciencia e Innovación and European ERDF under grants PID2021 123627OB-C51 and TED2021 130233B-C32. Marta Navarro is supported by Subvenciones para la contratación de personal investigador predoctoral by CIACIF/2021/413.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Main memory controller with multiple media technologies for big data workloads

  • Avargues, Miguel A.
  • Lurbe-Sempere, Manel
  • Yang, Rui
  • Zhu, Xiaoping
  • Wang, Guanhao
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
[EN] SRAM and DRAM memory technologies have been dominant in the implementations of memory subsystems. In recent years, and mainly driven by the huge memory demands of big data applications, NVRAM technology has emerged as a denser memory technology, enabling the design of new hybrid DRAM/NVRAM memory hierarchies that combine multiple memory media technologies to balance memory capacity, latency, cost, and endurance. Two main approaches are being applied to the design of hybrid memory hierarchies: the hybrid address space approach, which relies on the programmer or the operating system to choose the memory technology where each memory page should be stored; and the (only) NVM address space approach, where a faster technology (e.g. commodity DRAM) is needed to acts as a cache of NVRAM to boost the performance. This approach presents architectural challenges such the organization of metadata (e.g. cache tags) and the selection of the proper technology for each memory component. In contrast to existing approaches, this work proposes a memory controller that leverages novel memory technologies such as eDRAM and MRAM to mitigate NVRAM bus contention and improve the performance of the NVM address space. The devised solution proposes a two-level cache hierarchy in the memory controller: a SRAM sector cache and a (x)RAM cache. The (x)RAM cache, much denser, helps significantly reduce the number of accesses to NVRAM. Experimental results show that implementing the (x)RAM cache with eDRAM or MRAM is the best performing approach. Moreover, the eRAM is able to improve the SRAM cache miss penalty by up to 50% and 80%, and overall system performance by 15% and 23%., This work has been partially supported by Huawei under Agreement No: TC20210705020, by the Spanish Ministerio de Ciencia e Innovacion and European ERDF under grants PID2021-123627OB-C51 and TED2021-130233B-C32.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Advanced resource management: A hands-on master course in HPC and cloud computing

  • Pons-Escat, Lucía|||0000-0002-4582-7744
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
[EN] Resource management has become a major concern in dealing with performance and fairness in recent computing servers, including a wide variety of shared resources. To achieve high-performing and efficient systems, both hardware and software engineers must be thoroughly trained in effective resource management techniques. This paper introduces the GRE master course (Spanish acronym for Resource Management and Performance Evaluation in Cloud and High-Performance Workloads), which is being offered since Fall 2023. The course is taught by instructors with broad research expertise in resource management and performance evaluation. Subjects covered in this course include workload characterization, state-of-the-art resource management approaches, and performance evaluation tools and methodologies used in production systems. Management techniques are studied both in the context of HPC and cloud computing, where resource efficiency is becoming a primary concern. To enhance the learning experience, the course integrates theoretical concepts with a wide set of hands-on tasks carried out on recent real platforms. A real cloud virtualized environment is mimicked using typical software deployed in production systems such as Proxmox Virtual Environment. Students learn to use tools such as Linux Perf and Intel Vtune Profiler, which are commonly employed by researchers and practitioners to carry out typical tasks like performance bottleneck analysis from a microarchitectural perspective. Overall, the GRE course provides students with a solid foundation and skills in resource management by addressing current hot topics both in the industry and academia. Student satisfaction and learning outcomes prove the success of the GRE course and encourage us to continue in this direction., This work has been supported by the Spanish Ministerio de Ciencia e Innovación and European ERDF under grants PID2021-123627OB-C51 and TED2021-130233B-C32.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Dual Fast-Track Cache: Organizing Ring-Shaped Racetracks to Work as L1 Caches

  • Valero, Alejandro
  • Lorente Garcés, Vicente Jesús
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
[EN] Static Random-Access Memory (SRAM) is the fastest memory technology and has been the common design choice for implementing first-level (L1) caches in the processor pipeline, where speed is a key design issue that must be fulfilled. On the contrary, this technology offers much lower density compared to other technologies like Dynamic RAM, limiting L1 cache sizes of modern processors to a few tens of KB. This paper explores the use of slower but denser Domain Wall Memory (DWM) technology for L1 caches. This technology provides slow access times since it arranges multiple bits sequentially in a magnetic racetrack. To access these bits, they need to be shifted in order to place them under a header. A 1-bit shift usually takes one processor cycle, which can significantly hurt the application performance, making this working behavior inappropriate for L1 caches. Based on the locality (temporal and spatial) principles exploited by caches, this work proposes the Dual Fast-Track Cache (Dual FTC) design, a new approach to organizing a set of racetracks to build set-associative caches. Compared to a conventional SRAM cache, Dual FTC enhances storage capacity by 5x while incurring minimal shifting overhead, thereby rendering it a practical and appealing solution for L1 cache implementations. Experimental results show that the devised cache organization is as fast as an SRAM cache for 78% and 86% of the L1 data cache hits and L1 instruction cache hits, respectively (i.e., no shift is required). Consequently, due to the larger L1 cache capacities, significant system performance gains (by 22% on average) are obtained under the same silicon area., This work was supported in part by Agencia Estatal de Investigacion (AEI) under Grant PID2022-136454NB-C22/AEI/10.13039/501100011033; in part by the Spanish Ministerio de Ciencia e Innovacion and European ERDF under Grant PID2021-123627OB-C51 and Grant TED2021-130233B-C32; in part by European Union-NextGenerationE under Grant PDC2023-145851-00/MCIN/AEI/10.13039/501100011033; and in part by the Department of Science, University, and Knowledge Society, Government of Aragon under Grant gaZ: T58_23R research group.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Power, energy, and performance analysis of single-and multi-threaded applications in the ARM ThunderX2

  • Calero, Ibai
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
[EN] Energy efficiency has been a major concern in data centers, and the problem is exacerbated as its size continues to rise. However, the lack of tools to measure and handle this energy at a fine granularity (e.g., processor core or last-level cache) has translated into slow research advances in this topic. Understanding where (i.e., which components) and when (the point in time) energy consumption translates into minor performance improvements is of paramount importance to design any energy-aware scheduler. This paper characterizes the relationship between energy consumption and performance in a 28-core ARM ThunderX2 processor for both single-threaded and multi-threaded applications. This paper shows that single-threaded applications with high CPU activity maintain their performance in spite of the inter-application interference at shared resources, but this comes at the expense of higher power consumption. Conversely, applications that heavily utilize the L3 cache and memory consume less power but suffer significant performance degradation as interference levels rise. In contrast, multi-threaded applications show two distinct behaviors. On the one hand, some of them experience significant performance gains when they execute in a higher number of cores with more threads, which outweighs the increase in power consumption, leading to high energy efficiency., This work has been partially supported by the Spanish Ministerio de Ciencia e Innovacion and European ERDF under grants PID2021-123627OB-C51 and TED2021-130233B-C32.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

A modular approach to build a hardware testbed for cloud resource management research

  • Pons-Escat, Lucía|||0000-0002-4582-7744
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Pons Terol, Julio|||0000-0002-5654-6753
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
[EN] Research on resource management focuses on optimizing system performance and energy efficiency by distributing shared resources like processor cores, caches, and main memory among competing applications. This research spans a wide range of applications, including those from high-performance computing, machine learning, and mobile computing. Existing research frameworks often simplify research by concentrating on specific characteristics, such as the architecture of the computing nodes, resource monitoring, and representative workloads. For instance, this is typically the case with cloud systems, which introduce additional complexity regarding hardware and software requirements. To avoid this complexity during research, experimental frameworks are being developed. Nevertheless, proposed frameworks often fail regarding the types of nodes included, virtualization support, and management of critical shared resources. This paper presents Stratus, an experimental framework that overcomes these limitations. Stratus includes different types of nodes, a comprehensive virtualization stack, and the ability to partition the major shared resources of the system. Even though Stratus was originally conceived to perform cloud research, its modular design allows Stratus to be extended, broadening its research use on different computing domains and platforms, matching the complexity of modern cloud environments, as shown in the case studies presented in this paper., This work has been supported by the Spanish Ministerio de Universidades under the grant FPU18/01948 and by the Spanish Ministerio de Ciencia e Innovación and European ERDF under grants PID2021-123627OB-C51 and TED2021-130233B-C32, and by Generalitat Valenciana under Grant AICO/2021/266.