Pasar al contenido principal

MEJORA DE LA ARQUITECTURA DE SERVIDORES, SERVICIOS Y APLICACIONES

TIN2012-38341-C04-01

•
Nombre agencia financiadora Ministerio de Economía y Competitividad
Acrónimo agencia financiadora MINECO
Programa Programa Nacional de Proyectos de Investigación Fundamental
Subprograma Subprograma de Proyectos de Investigación Fundamental no Orientada
Convocatoria Proyectos I+D
Año convocatoria 2012
Plan Estatal 2008-2011 Plan Nacional de Investigación Científica, Desarrollo e Innovación Tecnológica
Unidad de gestión Secretaría de Estado de Investigación, Desarrollo e Innovación
Centro beneficiario UNIVERSITAT POLITÈCNICA DE VALÈNCIA
Identificador persistente http://dx.doi.org/10.13039/501100003329

Resultados relacionados

Resultados totales (Incluyendo duplicados): 30
Encontrada(s) 1 página(s)

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

Bringing Real Processorsto Labs

  • Gómez Requena, Crispín
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
The architecture of current processors has experienced great changes in the last years, leading to sophisticated multithreaded multicore processors. The inherent complexity of such processors makes difficult to update processor teaching to include current commercial products, especially at lab sessions where simplistic simulators are usually used. However, instructors are forced to reduce this gap if they want to properly prepare students in this topic. Dealing with these complex concepts at labs does not only help reinforce theoretical concepts but also has a positive effect in the students motivation. This article presents amethodology designed for the study of current microprocessor mechanisms in a gradual way without overwhelming students. The methodology is based on the use of a detailed simulation framework, used both in the academia and in the industry, which accurately models features from current processors. Due to the huge simulator complexity, it is introduced through several learning phases. Qualitative and quantitative results demonstrate that students are able to develop skills in a detailed simulator in a reasonable time period and, at the same time they learn the details of complex architectural mechanisms of commercial microprocessors., Contract grant sponsor: Spanish Government; Contract grant number: TIN2012-38341-C04-01
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Addressing bandwidth contention in SMT multicores through scheduling

  • Feliu-Pérez, Josué|||0000-0003-3017-4266
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Duato Marín, José Francisco
To mitigate the impact of bandwidth contention, which in some processes can yield to performance degradations up to 40%, we devise a scheduling algorithm that tackles main memory and L1 bandwidth contention. Experimental evaluation on a real system shows that the proposal achieves an average speedup by 5% with respect to Linux., This work was supported by the Spanish Ministerio de Economía y Competitividad (MINECO) and Plan E funds, under Grant TIN2012-38341-C04-01, and by the Intel Early Career Faculty Honor Program Award.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Speeding-up the fault-tolerance analysis of interconnection networks

  • Bermúdez Garzón, Diego Fernando
  • Gómez Requena, Crispín
  • López Rodríguez, Pedro Juan|||0000-0003-4544-955X
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
Analyzing the fault-tolerance of interconnection networks implies checking the connectivity of each sourcedestination pair. The size of the exploration space of such operation skyrockets with the network size and with the number of link faults. However, this problem is highly parallelizable since the exploration of each path between a source–destination pair is independent of the other paths. This paper presents an approach to analyze the fault-tolerance degree of multistage interconnection networks using GPUs in order to speed-up it. This approach uses CUDA as parallel programming tool on a GPU in order to take advantage of all available cores. Results show that the execution time of the fault-tolerance exploration can be significantly reduced., This work was supported by the Spanish Ministerio de Economía y Competitividad (MINECO) and by FEDER funds under Grant TIN2012-38341-C04-01.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Capítulo o parte de un libro (BookPart)

Enhancing Performance and Energy Consumption of HER Caches by Adding Associativity

  • Lorente Garcés, Vicente Jesús
  • Valero Bresó, Alejandro
  • Canal, Ramón
Unlike other previous techniques, the recently proposed Hard Error Recovery (HER) fault-tolerant cache provides 100% fault-coverage in L1 data caches. This full coverage makes the HER cache appropiate for fault-dominated future technology nodes. An n-way set-associative HER cache implements one cache way with fast SRAM banks and the remaining ways with eDRAM banks to address power and area. Since the number of eDRAM cache blocks used in a specific HER cache organization depends on the cache associativity (i.e., the implemented number of ways), we expect that the performance and energy consumption provided by a given HER cache design strongly depends on the cache geometry. In this work we study the behavior of the HER cache design when applied to a highly associative L1 cache like those found in some modern microprocessors. In particular this work explores a 32KB 8-way associative L1 data cache such as the one used in Intel Haswell microarchitecture. Experimental results show that, at low-power modes compared to a conventional cache with the same storage capacity and number of ways, area, leakage power, and dynamic energy savings of a 4-way HER cache are by 25%, 85%, and 62%, respectively. These percentages are further improved (by 40%, 89%, and 68%, respectively) when the cache associativity is increased to 8 ways, while the performance loss with respect to both an 8-way conventional cache and the 4-way HER cache is minimal., This work was supponed by Generalitat de Catalunya (200950R1250), by FP7 program of the European Commission (TRAMS-248789), by Spanish Ministerio de Economía y Competitividad (MINECO) and by FEDER funds under Grant TlN2012-38341-C04-01 and TIN2010-18368.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Cache-Hierarchy contention-aware scheduling in CMPs

  • Feliu-Pérez, Josué|||0000-0003-3017-4266
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • Duato Marín, José Francisco
To improve chip multiprocessor (CMP) performance, recent research has focused on scheduling strategies to mitigate main memory bandwidth contention. Nowadays, commercial CMPs implement multilevel cache hierarchies that are shared by several multithreaded cores. In this microprocessor design, contention points may appear along the whole memory hierarchy. Moreover, this problem is expected to aggravate in future technologies, since the number of cores and hardware threads, and consequently the size of the shared caches increase with each microprocessor generation. This paper characterizes the impact on performance of the different contention points that appear along the memory subsystem. The analysis shows that some benchmarks are more sensitive to contention in higher levels of the memory hierarchy (e.g., shared L2) than to main memory contention. In this paper, we propose two generic scheduling strategies for CMPs. The first strategy takes into account the available bandwidth at each level of the cache hierarchy. The strategy selects the processes to be coscheduled and allocates them to cores to minimize contention effects. The second strategy also considers the performance degradation each process suffers due to contention-aware scheduling. Both proposals have been implemented and evaluated in a commercial single-threaded quad-core processor with a relatively small two-level cache hierarchy. The proposals reach, on average, a performance improvement by 5.38 and 6.64 percent when compared with the Linux scheduler, while this improvement is by 3.61 percent for an state-of-the-art memory contention-aware scheduler under the evaluated mixes., This work was supported by the Spanish MINECO under Grant TIN2012-38341-C04-01, and by the Universitat Politecnica de Valencia under Grant PAID-05-12 SP20120748.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Analyzing the Optimal Voltage/Frequency Pair in Fault-Tolerant Caches

  • Lorente Garcés, Vicente Jesús
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • Valero, Alejandro
  • Foglia, Pierfrancesco
[EN] When the processor works at very-low voltages to save energy, failures in SRAM cells increase exponentially at voltages below VCCmin. In this context, current SRAM-error detection and correction proposals incur on a significant performance penalty since they increase access latency and disable cache lines that cannot be corrected, so decreasing the effective cache capacity. This reduction implies more cache misses, so enlarging the execution time which, contrary to expected, can turn in higher energy consumption. This paper characterizes SRAM failures at very-low voltages and presents an evaluation methodology to analyze the impact on energy consumption of error correction approaches. To do so, several voltage/frequency pairs are studied and the optimal pair is identified from an energy point of view. To focus the research, experimental results have been obtained for the recently proposed fault-tolerant HER cache. Results show that, for a 32nm technology node, the voltage/frequency pair of 0.45V/800MHz, which induces by 31% SRAM failure rate, provides the lowest overall energy consumption (by 62% energy savings compared to a non-faulty conventional cache)., This work was supported by the Spanish Ministerio de Economia y Competitividad (MINECO) and FEDER funds under Grant TIN2012-38341-C04-01. Additionally, it was also supported by the Intel Early Career Honor Programme Award and the Intel Doctoral Student Honor Programme Award.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

A complete and efficient CUDA-sharing solution for HPC clusters

  • Peña Monferrer, Antonio José
  • Reaño González, Carlos
  • Mayo Gual, Rafael
  • Quintana-Orti, Enrique S.
  • Duato Marín, José Francisco
  • Silla, Federico|||0000-0002-6435-1200
In this paper we detail the key features, architectural design, and implementation of rCUDA, an advanced framework to enable remote and transparent GPGPU acceleration in HPC clusters. rCUDA allows decoupling GPUs from nodes, forming pools of shared accelerators, which brings enhanced flexibility to cluster configurations. This opens the door to configurations with fewer accelerators than nodes, as well as permits a single node to exploit the whole set of GPUs installed in the cluster. In our proposal, CUDA applications can seamlessly interact with any GPU in the cluster, independently of its physical location. Thus, GPUs can be either distributed among compute nodes or concentrated in dedicated GPGPU servers, depending on the cluster administrator’s policy. This proposal leads to savings not only in space but also in energy, acquisition, and maintenance costs. The performance evaluation in this paper with a series of benchmarks and a production application clearly demonstrates the viability of this proposal. Concretely, experiments with the matrix–matrix product reveal excellent performance compared with regular executions on the local GPU; on a much more complex application, the GPU-accelerated LAMMPS, we attain up to 11x speedup employing 8 remote accelerators from a single node with respect to a 12-core CPU-only execution. GPGPU service interaction in compute nodes, remote acceleration in dedicated GPGPU servers, and data transfer performance of similar GPU virtualization frameworks are also evaluated. 2014 Elsevier B.V. All rights reserved., This work was supported by the Spanish Ministerio de Economia y Competitividad (MINECO) and by FEDER funds under Grant TIN2012-38341-004-01. It was also supported by MINECO, FEDER funds, under Grant TIN2011-23283, and by the Fundacion Caixa-Castello Bancaixa, Grant P11B2013-21. This work was also supported in part by the U.S. Department of Energy, Office of Science, under contract DE-AC02-06CH11357. Authors are grateful for the generous support provided by Mellanox Technologies to the rCUDA Project. The authors would also like to thank Adrian Castello, member of The rCUDA Development Team, for his hard work on rCUDA.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

A Research-Oriented Course on Advanced Multicore Architecture

  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
  • Selfa Oliver, Vicent
Multicore processors have become ubiquitous in our real life in devices like smartphones, tablets, etc. In fact, they are present in almost all segments of the computing market, from supercomputers to embedded devices. The huge market competence have lead industry and academia to develop vertiginous technological and architectural advances. The fast evolution that are still experiencing current multicores makes difficult for instructors to offer computer architecture courses with updated contents, preferably showing the industry and academia research trends. To deal with this shortcoming, authors consider that a research-oriented course is the most appropriate solution. This paper presents an advanced computer architecture course called Advanced Multicore Architectures, offered in 2015. The course covers the basic topics of multicore architecture and has been organized in four main modules regarding multicore basis, performance evaluation, advanced caching, and main memory organization. The course follows a research-oriented approach that covers theoretical concepts at lectures in which recent research papers are analyzed to provide students a wide view of current trends. Moreover, additional teaching methods like lab sessions with a state-of-the-art multicore simulator or research-oriented exercises have been used with the aim of introducing students to research in these topics. To achieve this fully research-oriented methodology, about 40% of the time is devoted to labs and exercises., This work was supported by the Spanish Ministerio de Economía y Competitividad (MINECO) and by FEDER funds under Grant TIN2012-38341-C04-01, and by the Intel Early Career Faculty Honor Program Award.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Power consumption management in fat-tree interconnection networks

  • Alonso Díaz, Marina|||0000-0001-7079-5898
  • Coll, Salvador|||0000-0002-7795-7252
  • Martínez-Rubio, Juan-Miguel|||0000-0002-9748-0332
  • López Rodríguez, Pedro Juan|||0000-0003-4544-955X
  • Santonja Gisbert, Vicente
As higher communication bandwidth is required in current designs of high performance parallel computers, the amount of power consumed by the interconnection network also increases. Fat-tree is one of the most popular topologies in high performance interconnection networks aiming at low latency, efficient collective communication and scalability. We present a new methodology for managing power consumption of fat-tree interconnection networks. Our proposal is based on dynamically adjusting available network bandwidth according to traffic requirements. To meet this goal, we define a mechanism for managing the operating status of network links as a function of network load. Our main contributions include a complete definition of the mechanism and a tuning methodology based on its sensitivity and aggressiveness in terms of potential power savings. Results show that our proposal can provide significant power savings (up to 67% in a 4-ary 4-tree) with no changes in the underlying routing algorithm, with minimal impact on network performance. Experiments conducted on a 16-ary 3-tree topology provide up to 36% energy savings with performance degradation below 1.10%., This work was supported by the Spanish Ministerio de Economa y Competitividad (MINECO) and by FEDER funds under Grant TIN2012-38341-C04-01
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Exploiting Reuse Information to Reduce Refresh Energy in On-Chip eDRAM Caches

  • Valero Bresó, Alejandro
  • Duato Marín, José Francisco
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
This work introduces a novel refresh mechanism that leverages reuse information to decide which blocks should be refreshed in an energy-aware eDRAM last-level cache. Experimental results show that, compared to a conventional eDRAM cache, the energy-aware approach achieves refresh energy savings up to 71%, while the reduction on the overall dynamic energy is by 65% with negligible performance losses., This work was supported by the Spanish Ministerio de Economía y Competitividad (MINECO) and Plan E funds, under Grants TIN-2009-14475-C04-01 and TIN2012-38341-C04-01.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

L1-Bandwidth Aware Thread Allocation in Multicore SMT Processors

  • Feliu-Pérez, Josué|||0000-0003-3017-4266
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Duato Marín, José Francisco
Improving the utilization of shared resources is a key issue to increase performance in SMT processors. Recent work has focused on resource sharing policies to enhance the processor performance, but their proposals mainly concentrate on novel hardware mechanisms that adapt to the dynamic resource requirements of the running threads. This work addresses the L1 cache bandwidth problem in SMT processors experimentally on real hardware. Unlike previous work, this paper concentrates on thread allocation, by selecting the proper pair of co-runners to be launched to the same core. The relation between L1 bandwidth requirements of each benchmark and its performance (IPC) is analyzed. We found that for individual benchmarks, performance is strongly connected to L1 bandwidth consumption, and this observation remains valid when several co-runners are launched to the same SMT core. Based on these findings we propose two L1 bandwidth aware thread to core (t2c) allocation policies, namely Static and Dynamic t2c allocation, respectively. The aim of these policies is to properly balance L1 bandwidth requirements of the running threads among the processor cores. Experiments on a Xeon E5645 processor show that the proposed policies significantly improve the performance of the Linux OS kernel regardless the number of cores considered., This work was supported by the Spanish Ministerio de Econom´ıa y Competitividad (MINECO) and by FEDER funds under Grant TIN2012-38341-C04-01; and by Programa de Apoyo a la Investigacion y Desarrollo (PAID-05-12) of the ´ Universitat Politecnica de Val ` encia under Grant SP20120748
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

A reuse-based refresh policy for energy-aware eDRAM caches

  • Valero Bresó, Alejandro
  • Kaeli, David R.
  • Duato Marín, José Francisco
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
DRAM technology requires refresh operations to be performed in order to avoid data loss due to capacitance leakage. Refresh operations consume a significant amount of dynamic energy, which increases with the storage capacity. To reduce this amount of energy, prior work has focused on reducing refreshes in off-chip memories. However, this problem also appears in on-chip eDRAM memories implemented in current low-level caches. The refresh energy can dominate the dynamic consumption when a high percentage of the chip area is devoted to eDRAM cache structures. Replacement algorithms for high-associativity low-level caches select the victim block avoiding blocks more likely to be reused soon. This paper combines the state-of-the-art MRUT replacement algorithm with a novel refresh policy. Refresh operations are performed based on information produced by the replacement algorithm. The proposed refresh policy is implemented on top of an energy-aware eDRAM cache architecture, which implements bank-prediction and swap operations to save energy. Experimental results show that, compared to a conventional eDRAM design, the proposed energyaware cache can achieve by 72% refresh energy savings. Considering the entire on-chip memory hierarchy consumption, the overall energy savings are 30%. These benefits come with minimal impact on performance (by 1.2%) and area overhead (by 0.4%). 2014 Elsevier B.V. All rights reserved., This work was supported by the Spanish Ministerio de Economia y Competitiyidad (MINECO) and FEDER funds under Grant TIN2012-38341-C04-01. Additionally, it was also supported by the Intel Early Career Honor Programme Award and the Intel Doctoral Student Honor Programme Award. Prof. David R. Kaeli was supported in part by an NSF Award, CNS-1319501.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

On the design of a demo for exhibiting rCUDA

  • Reaño González, Carlos
  • Pérez López, Ferran
  • Silla, Federico|||0000-0002-6435-1200
CUDA is a technology developed by NVIDIA which provides a parallel computing platform and programming model for NVIDIA GPUs and compatible ones. It takes benefit from the enormous parallel processing power of GPUs in order to accelerate a wide range of applications, thus reducing their execution time. rCUDA (remote CUDA) is a middleware which grants applications concurrent access to CUDA-compatible devices installed in other nodes of the cluster in a transparent way so that applications are not aware of accessing a remote device. In this paper we present a demo which shows, in real time, the overhead introduced by rCUDA in comparison to CUDA when running image filtering applications. The approach followed in this work is to develop a graphical demo which contains both an appealing design and technical contents., This work was funded by the Spanish MINECO and FEDER funds under Grant TIN2012-38341-C04-01. Authors are also grateful for the generous support provided by Mellanox Technologies and the equipment donated by NVIDIA Corporation
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Combining RAM technologies for hard-error recovery in L1 data caches working at very-low power modes

  • Lorente Garcés, Vicente Jesús
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • López Rodríguez, Pedro Juan|||0000-0003-4544-955X
  • Valero Bresó, Alejandro
  • Canal, Ramón
  • Duato Marín, José Francisco
Low-power modes in modern microprocessors rely on low frequencies and low voltages to reduce the energy budget. Nevertheless, manufacturing induced parameter variations can make SRAM cells unreliable producing hard errors at supply voltages below Vccmin. Recent proposals provide a rather low fault-coverage due to the fault coverage/overhead trade-off. We propose a new faulttolerant L1 cache, which combines SRAM and eDRAM cells in L1 data caches to provide 100% SRAM hard-error fault coverage. Results show that, compared to a conventional cache and assuming 50% failure probability at low-power mode, leakage and dynamic energy savings are by 85% and 62%, respectively, with a minimal impact on performance., This work was supported by the Spanish MICINN (TIN2010-18368) with the Consolider-Ingenio 2010 Programme co-funded by the European Commission FEDER funds (CSD2006-00046) and co-funded with the Plan E funds (TIN2009-14475-C04-01). Additionaly, it was supported by Generalitat de Catalunya (2009SGR1250), by FP7 program of the European Commission (TRAMS-248789), and by Spanish MINECO (TIN2012-38341-C04-01).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Efficient TLB-Based Detection of Private Pages in Chip Multiprocessors

  • Esteve García, Albert
  • Ros Bardisa, Alberto
  • Robles Martínez, Antonio
  • Duato Marín, José Francisco
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
Most of the data referenced by sequential and parallel applications running in current chip multiprocessors are referenced by a single thread, i.e., private. Recent proposals leverage this observation to improve many aspects of chip multiprocessors, such as reducing coherence overhead or the access latency to distributed caches. The effectiveness of those proposals depends to a large extent on the amount of detected private data. However, the mechanisms proposed so far do not consider neither thread migration nor the private use of data within different application phases. As a result, a considerable amount of private data is not detected. In order to increase the detection of private data, we propose a TLB-based mechanism that is able to account for both thread migration and application phases. Simulation results show that the average number of pages detected as private significantly increases from 43 percent in previous proposals up to 79 percent in ours while keeping a reasonable TLB miss rate. Furthermore, when our proposal is used to deactivate the coherence for private data in a directory protocol, it improves execution time by 13.5 percent, on average, with respect to previous techniques., This work was jointly supported by the MINECO and European Commission (FEDER funds) under the project TIN2012-38341-C04-01/03 and the Fundacion Seneca-Agencia de Ciencia y Tecnologia de la Region de Murcia under the project Jovenes Lideres en Investigacion 18956/JLI/13. Albert Esteve is the corresponding author.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

PS directory: a scalable multilevel directory cache for CMPs

  • Valls, Joan J
  • Ros Bardisa, Alberto
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
As the number of cores increases in current and future chip-multiprocessor (CMP) generations, coherence protocols must rely on novel hardware structures to scale in terms of performance, power, and area. Systems that use directory information for coherence purposes are currently the most scalable alternative. This paper studies the important differences between the directory behavior of private and shared blocks, which claim for a separate management of both types of blocks at the directory. We propose the PS directory, a two-level directory cache that keeps the reduced number of frequently accessed shared entries in a small and fast first-level cache, namely Shared cache, and uses a larger and slower second-level Private cache to track the large amount of private blocks. Entries in the Private cache do not implement the sharer vector, which allows important silicon area savings. Speed and area reasons suggest the use of eDRAM technology, much denser but slower than SRAM technology, for the Private cache, which in turn brings energy savings. Experimental results for a 16-core CMP show that, compared to a conventional directory, the PS directory improves performance by 14 % while reducing silicon area and energy consumption by 34 and 27 %, respectively. Also, compared to the state-of-the-art Multi-Grain Directory, the PS directory apart from increasing performance, it reduces power by 18.7 %, and provides more scalability in terms of area., This work has been jointly supported by the MINECO and European Commission (FEDER funds) under the project TIN2012-38341-C04-01 and the Fundacion Seneca-Agencia de Ciencia y Tecnologia de la Region de Murcia under the project Jovenes Lideres en Investigacion 18956/JLI/13.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

A Family of Fault-Tolerant Efficient Indirect Topologies

  • Bermúdez Garzón, Diego Fernando
  • Gómez Requena, Crispín
  • Duato Marín, José Francisco
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
  • López Rodríguez, Pedro Juan|||0000-0003-4544-955X
On the one hand, performance and fault-tolerance of interconnection networks are key design issues for high performance computing (HPC) systems. On the other hand, cost should be also considered. Indirect topologies are often chosen in the design of HPC systems. Among them, the most commonly used topology is the fat-tree. In this work, we focus on getting the maximum benefits from the network resources by designing a simple indirect topology with very good performance and fault-tolerance properties, while keeping the hardware cost as low as possible. To do that, we propose some extensions to the fat-tree topology to take full advantage of the hardware resources consumed by the topology. In particular, we propose three new topologies with different properties in terms of cost, performance and fault-tolerance. All of them are able to achieve a similar or better performance results than the fat-tree, providing also a good level of fault-tolerance and, contrary to most of the available topologies, these proposals are able to tolerate also faults in the links that connect to end nodes., This work was supported by the Spanish Ministerio de Economia y Competitividad (MINECO) and by FEDER funds under Grant TIN2012-38341-C04-01.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Tuning remote GPU virtualization for InfiniBand networks

  • Reaño González, Carlos
  • Silla, Federico|||0000-0002-6435-1200
In the past few years, a tendency towards using InfiniBand networks to interconnect high performance computing clusters can be observed. Thus, most of the supercomputers appearing in the TOP500 list either use Ethernet or InfiniBand interconnects. Regarding the latter, the complexity of the InfiniBand programming API (i.e., InfiniBand Verbs) makes it difficult for applications to get the maximum performance of these networks. In this paper we expose how we have tuned a remote GPU virtualization framework whose communications module is implemented using InfiniBand Verbs. The net result is a noticeable increase in the performance of this framework, significantly reducing the gap between remote and local GPUs., This work was funded by the Spanish MINECO and FEDER funds under Grant TIN2012-38341-C04-01. Authors are also grateful for the generous support provided by Mellanox Technologies.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

A dynamic execution time estimation model to save energy in heterogeneous multicores running periodic tasks

  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • Hassan Mohamed, Houcine|||0000-0002-2038-0587
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • March Cabrelles, José Luis
  • Duato Marín, José Francisco
Nowadays, real-time embedded applications have to cope with an increasing demand of functionalities, which require increasing processing capabilities. With this aim real-time systems are being implemented on top of high-performance multicore processors that run multithreaded periodic workloads by allocating threads to individual cores. In addition, to improve both performance and energy savings, the industry is introducing new multicore designs such as ARM’s big.LITTLE that include heterogeneous cores in the same package. A key issue to improve energy savings in multicore embedded real-time systems and reduce the number of deadline misses is to accurately estimate the execution time of the tasks considering the supported processor frequencies. Two main aspects make this estimation difficult. First, the running threads compete among them for shared resources. Second, almost all current microprocessors implement Dynamic Voltage and Frequency Scaling (DVFS) regulators to dynamically adjust the voltage/frequency at run-time according to the workload behavior. Existing execution time estimation models rely on off-line analysis or on the assumption that the task execution time scales linearly with the processor frequency, which can bring important deviations since the memory system uses a different power supply. In contrast, this paper proposes the Processor–Memory (Proc–Mem) model, which dynamically predicts the distinct task execution times depending on the implemented processor frequencies. A power-aware EDF (Earliest Deadline First)-based scheduler using the Proc–Mem approach has been evaluated and compared against the same scheduler using a typical Constant Memory Access Time model, namely CMAT. Results on a heterogeneous multicore processor show that the average deviation of Proc–Mem is only by 5.55% with respect to the actual measured execution time, while the average deviation of the CMAT model is 36.42%. These results turn in important energy savings, by 18% on average and up to 31% in some mixes, in comparison to CMAT for a similar number of deadline misses. © 2015 Elsevier B.V. All rights reserved., This work was supported by the Spanish Ministerio de Economia y Competitividad (MINECO) and by FEDER funds under Grant TIN2012-38341-004-01, and by the Intel Early Career Faculty Honor Program Award.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Embedded GPU and multicore processors for emotional-based mobile robotic agents

  • Francisco Almenar
  • Domínguez Montagud, Carlos Pascual
  • Hassan Mohamed, Houcine|||0000-0002-2038-0587
  • Martínez-Rubio, Juan-Miguel|||0000-0002-9748-0332
  • López Rodríguez, Pedro Juan|||0000-0003-4544-955X
Control architectures based on emotions are becoming promising solutions for the implementation of future robotic systems. The basic controllers of this architecture are the emotional processes that decide which behaviors the robot must activate to fulfill the objectives. The number of emotional processes increases (hundreds of millions/s) with the complexity level of the application, limiting the processing capacity of a main processor to solve the complex problems. Fortunately, the potential parallelism of emotional processes permits their execution in parallel, hence enabling the computing power to tackle the complex dynamic problems. In this paper, Graphic Processing Unit (GPU), multicore processors and single instruction multiple data (SIMD) instructions are used to provide parallelism for the emotional processes. Different GPUs, multicore processors and SIMD instruction sets are evaluated and compared to analyze their suitability to cope with robotic applications. The applications are set-up taking into account different environmental conditions, robot dynamics and emotional states. Experimental results show that, despite the fact that GPUs have a bottleneck in the data transmission between the host and the device, the evaluated GTX 670 GPU provides a performance of more than one order of magnitude greater than the initial implementation of the architecture on a single core. Thus, all complex proposed application problems can be solved using the GPU technology in contrast to the first prototype where only 55% of them could be solved. Using AVX SIMD instructions, the performance of the architecture is increased in 3.25 times in relation to the first implementation. Thus, from the 27 proposed applications about 88.8% are solved. In the case of the SSE SIMD instructions, the performance is almost doubled and the robot could solve about 74% of the proposed application problems. The use of AVX and SSE SIMD instructions provides almost the same performance as a quad- and a dual-core, respectively, with the advantage that they do not add any additional hardware cost., This work was supported by the Spanish Ministerio de Economia y Competitividad (MINECO) and by FEDER funds under Grant TIN2012-38341-C04-01 and by Universitat Politecnica de Valencia under Grant PAID/2012/325.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

The k-ary n-direct s-indirect family of topologies for large-scale interconnection networks

  • Peñaranda Cebrián, Roberto
  • Gómez Requena, Crispín
  • Duato Marín, José Francisco
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
  • López Rodríguez, Pedro Juan|||0000-0003-4544-955X
In large-scale supercomputers, the interconnection network plays a key role in system performance. Network topology highly defines the performance and cost of the interconnection network. Direct topologies are sometimes used due to its reduced hardware cost, but the number of network dimensions is limited by the physical 3D space, which leads to an increase of the communication latency and a reduction of network throughput for large machines. Indirect topologies can provide better performance for large machines, but at higher hardware cost. In this paper, we propose a new family of hybrid topologies, the k-ary n-direct s-indirect, that combines the best features from both direct and indirect topologies to efficiently connect an extremely high number of processing nodes. The proposed network is an n-dimensional topology where the k nodes of each dimension are connected through a small indirect topology of s stages. This combination results in a family of topologies that provides high performance, with latency and throughput figures of merit close to indirect topologies, but at a lower hardware cost. In particular, it doubles the throughput obtained per cost unit compared with indirect topologies in most of the cases. Moreover, their fault-tolerance degree is similar to the one achieved by direct topologies built with switches with the same number of ports., This work was supported by the Spanish Ministerio de Economa y Competitividad (MINECO) and by FEDER funds under Grant TIN2012-38341-C04-01 and by Programa de Ayudas de Investigacion y Desarrollo (PAID) from Universitat Politecnica de Valencia.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Efficient register renaming and recovery for high-performance processors

  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • López Rodríguez, Pedro Juan|||0000-0003-4544-955X
  • Ubal Tena, Rafael
Modern superscalar processors implement register renaming using either random access memory (RAM) or content-addressable memories (CAM) tables. The design of these structures should address both access time and misprediction recovery penalty. Although direct-mapped RAMs provide faster access times, CAMs are more appropriate to avoid recovery penalties. The presence of associative ports in CAMs, however, prevents them from scaling with the number of physical registers and pipeline width, negatively impacting performance, area, and energy consumption at the rename stage. In this paper, we present a new hybrid RAM CAM register renaming scheme, which combines the best of both approaches. In a steady state, a RAM provides fast and energy-efficient access to register mappings. On misspeculation, a low-complexity CAM enables immediate recovery. Experimental results show that in a four-way state-ofthe- art superscalar processor, the new approach provides almost the same performance as an ideal CAM-based renaming scheme, while dissipating only between 17% and 26% of the original energy and, in some cases, consuming less energy than purely RAM-based renaming schemes. Overall, the silicon area required to implement the hybrid RAM CAM scheme does not exceed the area required by conventional renaming mechanisms., This work was supported in part by the Spanish MINECO under Grant TIN2012-38341-C04-01.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Enhancing the L1 Data Cache Design to Mitigate HCI

  • Valero Bresó, Alejandro
  • Miralaei, Negar
  • Jones, Timothy M.
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
[EN] Over the lifetime of a microprocessor, the Hot Carrier Injection (HCI) phenomenon degrades the threshold voltage, which causes slower transistor switching and eventually results in timing violations and faulty operation. This effect appears when the memory cell contents flip from logic '0' to '1' and vice versa. In caches, the majority of cell flips are concentrated into only a few of the total memory cells that make up each data word. In addition, other researchers have noted that zero is the most commonly-stored data value in a cache, and have taken advantage of this behavior to propose data compression and power reduction techniques. Contrary to these works, we use this information to extend the lifetime of the caches by introducing two microarchitectural techniques that spread and reduce the number of flips across the first-level (L1) data cache cells. Experimental results show that, compared to the conventional approach, the proposed mechanisms reduce the highest cell flip peak up to 65.8 percent, whereas the threshold voltage degradation savings range from 32.0 to 79.9 percent depending on the application., This work has been supported by the Spanish Ministerio de Economia y Competitividad (MINECO), by FEDER funds through Grant TIN2012-38341-C04-01, by the Intel Early Career Faculty Honor Program Award, by a HiPEAC Collaboration Grant funded by the FP7 HiPEAC Network of Excellence under grant agreement 287759, and by the Engineering and Physical Sciences Research Council (EPSRC) through Grants EP/K026399/1 and EP/J016284/1. Additional data related to this publication are available in the data repository at https://www.repository.cam.ac.uk/handle/1810/249006.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Bandwidth-Aware On-Line Scheduling in SMT Multicores

  • Feliu-Pérez, Josué|||0000-0003-3017-4266
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Duato Marín, José Francisco
The memory hierarchy plays a critical role on the performance of current chip multiprocessors. Main memory is shared by all the running processes, which can cause important bandwidth contention. In addition, when the processor implements SMT cores, the L1 bandwidth becomes shared among the threads running on each core. In such a case, bandwidth-aware schedulers emerge as an interesting approach to mitigate the contention. This work investigates the performance degradation that the processes suffer due to memory bandwidth constraints. Experiments show that main memory and L1 bandwidth contention negatively impact the process performance; in both cases, performance degradation can grow up to 40 percent for some of applications. To deal with contention, we devise a scheduling algorithm that consists of two policies guided by the bandwidth consumption gathered at runtime. The process selection policy balances the number of memory requests over the execution time to address main memory bandwidth contention. The process allocation policy tackles L1 bandwidth contention by balancing the L1 accesses among the L1 caches. The proposal is evaluated on a Xeon E5645 platform using a wide set of multiprogrammed workloads, achieving performance benefits up to 6.7 percent with respect to the Linux scheduler., This work was supported by the Spanish Ministerio de Economia y Competitividad (MINECO) and by FEDER funds under Grant TIN2012-38341-C04-01, and by the Intel Early Career Faculty Honor Program Award.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Impact of partitioning cache schemes on the cache hierarchy of SMT processors

  • Kenyon, Samantha
  • López, Sonia
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
Power consumption is becoming an increasingly important component of processor design. As technology node shrinks both static and dynamic power become more relevant. This is particularly critical for the cache hierarchy. Previous implementations mainly focus on reducing only one kind of power in the cache, either static or dynamic. However, for a more robust approach that will remain relevant as technology continues to shrink, both aspects of power need to be addressed. Recent processors, e.g. Intel Core or IBM Power8, implement simultaneous multithreading (SMT) cores to hide high memory latencies. In these systems, the dynamic energy in the L1 cache is even more stressed since this cache level is shared by several threads running on the same core. This paper proposes and evaluates the use of phase adaptive caches in all structures of a 3-level cache hierarchy of a SMT cores. Compared to the use of conventional caches, our work results on significant dynamic and leakage energy savings with scarce performance impact., This work was supported by the Spanish Ministerio de Economía y Competitividad (MINECO) and by FEDER funds under Grant TIN2012–38341–C04–01.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Addressing fairness in SMT multicores with a progress-aware scheduler

  • Feliu-Pérez, Josué|||0000-0003-3017-4266
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Duato Marín, José Francisco
Current SMT (simultaneous multithreading) processors co-schedule jobs on the same core, thus sharing core resources like L1 caches. In SMT multicores, threads also compete among themselves for uncore resources like the LLC (last level cache) and DRAM modules. Per process performance degradation over isolated execution mainly depends on process resource requirements and the resource contention induced by co-runners. Consequently, the running processes progress at different pace. If schedulers are not progress aware, the unpredictable execution time caused by unfairness can introduce undesirable behaviors on the system such as difficulties to keep priority-based scheduling. This work proposes a job scheduler for SMT multicores that provides fairness to the execution of multiprogrammed workloads. To this end, the scheduler estimates per-process standalone performance by periodically creating low-contention co-schedules. These estimates are used to compute the per process progress. Then, those processes with less progress are prioritized to enhance fairness. Experimental results on a Intel Xeon with six dual-threaded SMT cores show that the proposed scheduler reduces unfairness, on average, by 3× over Linux OS. Moreover, thanks to the tread to core allocation policy, the scheduler slightly improves throughput and turnaround time., This work was supported by the Spanish Ministerio de Econom´ıa y Competitividad (MINECO) and Plan E funds, under Grant TIN2012-38341-C04-01, and by the Intel Early Career Faculty Honor Program Award
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Capítulo o parte de un libro (BookPart)

Temporal-Aware Mechanism to Detect Private Data in Chip Multiprocessors

  • Ros Bardisa, Alberto
  • Cuesta Sáez, Blas Antonio
  • Robles Martínez, Antonio
  • Duato Marín, José Francisco
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
Most of the data referenced by sequential and parallel applications running in current chip multiprocessors are referenced by only one thread and can be considered as private data. A lot of recent proposals leverage this observation to improve many aspects of chip multiprocessors, such as reducing coherence overhead or the access latency to distributed caches. The effectiveness of those proposals depend to a large extent on the amount of detected private data. However, the mechanisms proposed so far do not consider thread migration and the private use of data within different application phases. As a result, a considerable amount of data is not detected as private. In order to make this detection more accurate and reaching more significant improvements, we propose a mechanism that is able to account for both thread migration and private data within application phases. Simulation results for 16-core systems show that, thanks to our mechanism, the average number of pages detected as private significantly increases from 43% in previous proposals up to 74% in ours. Finally, when our detection mechanism is used to deactivate the coherence for private data in a directory protocol, our proposal improves execution time by 13% with respect to previous proposals., This work was supported by the Spanish MINECO, as well as European Commission FEDER funds, under grant TIN2012-38341-C04-01/03 and by the VIRTICAL project (grant agreement no 288574) which is funded by the European Commission within the Research Programme FP7.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Design of Hybrid Second-Level Caches

  • Valero Bresó, Alejandro
  • Duato Marín, José Francisco
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • López Rodríguez, Pedro Juan|||0000-0003-4544-955X
In recent years, embedded dynamic random-access memory (eDRAM) technology has been implemented in last-level caches due to its low leakage energy consumption and high density. However, the fact that eDRAM presents slower access time than static RAM (SRAM) technology has prevented its inclusion in higher levels of the cache hierarchy. This paper proposes to mingle SRAM and eDRAM banks within the data array of second-level (L2) caches. The main goal is to achieve the best trade-off among performance, energy, and area. To this end, two main directions have been followed. First, this paper explores the optimal percentage of banks for each technology. Second, the cache controller is redesigned to deal with performance and energy. Performance is addressed by keeping the most likely accessed blocks in fast SRAM banks. In addition, energy savings are further enhanced by avoiding unnecessary destructive reads of eDRAM blocks. Experimental results show that, compared to a conventional SRAM L2 cache, a hybrid approach requiring similar or even lower area speedups the performance on average by 5.9 percent, while the total energy savings are by 32 percent. For a 45 nm technology node, the energy-delay-area product confirms that a hybrid cache is a better design than the conventional SRAM cache regardless of the number of eDRAM banks, and also better than a conventional eDRAM cache when the number of SRAM banks is an eighth of the total number of cache banks, This work was supported by the Spanish Ministerio de Economia y Competitividad (MINECO) and FEDER funds under Grant TIN2012-38341-C04-01. Additionally, it was also supported by the Intel Early Career Honor Programme Award and the Intel Doctoral Student Honor Programme Award. Part of this work was done while Alejandro Valero was a visiting researcher with the Northeastern University Computer Architecture Research laboratory conducted by Prof. David R. Kaeli, Department of Electrical and Computer Engineering, Northeastern University, Boston, MA, USA.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

PS-Cache: an energy-efficient cache design for chip multiprocessors

  • Valls, Joan J
  • Ros Bardisa, Alberto
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
Power consumption has become a major design concern in current high-performance chip multiprocessors, and this problem exacerbates with the number of core counts. A significant fraction of the total power budget is often consumed by on-chip caches, thus important research has focused on reducing energy consumption in these structures. To enhance performance, on-chip caches are being deployed with a high associativity degree. Consequently, accessing concurrently all the ways in the cache set is costly in terms of energy. This paper presents the PS-Cache architecture, an energy-efficient cache design that reduces the number of accessed ways without hurting the performance. The PS-Cache takes advantage of the private-shared knowledge of the referenced block to reduce energy by accessing only those ways holding the kind of block looked up. Experimental results show that, on average, the PS-Cache architecture can reduce the dynamic energy consumption of L1 and L2 caches by 22 and 40%, respectively., This work has been jointly supported by the MINECO and European Commission (FEDER funds) under the project TIN2012-38341-C04-01 and the Fundaci’on Seneca-Agencia de Ciencia y Tecnología de la Región de Murcia under the project Jóvenes Líderes en Investigación 18956/JLI/13.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Current challenges in simulations of HPC systems

  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
Simulation of many-core HPC systems is nowadays an active and fruitful area of research. Recent and future proposals are driven by the need of a fast, efficient, and comprehensive simulation framework. This simulation framework should be complete in several ways. First, it should model a wide range of components and provide the mechanisms necessary to plug-in more components as needed. Second, it should allow the designer to focus on critical components while avoiding a large part of the simulation complexity. Each of these components should be able to be evaluated with multiple models with distinct detail levels, ranging from simply analytical models to detailed cycle-accurate simulations. Third, a complete simulation framework should provide a wide range of metrics of interest for the designer and the market. Finally, support for heterogeneous architectures combining CPU and GPU, as well as some degree of reconfigurability is surely required. Building such titanic framework is and will be a collaborative process between researchers around the globe and it is expected to be a hot research topic for the next years., This work has been supported by the Spanish Ministerio de Economía y Competitividad (MINECO), by FEDER funds through Grant TIN2012-38341-C04-01, and by the Intel Early Career Faculty Honor Program Award.