Skip to main content

Resultados relacionados

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

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

Efficient Management of Cache Accesses to Boost GPGPU Memory Subsystem Performance

  • Candel-Margaix, Francisco
  • Valero Bresó, Alejandro
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
[EN] To support the massive amount of memory accesses that GPGPU applications generate, GPU memory hierarchies are becoming more and more complex, and the Last Level Cache (LLC) size considerably increases each GPU generation. This paper shows that counter-intuitively, enlarging the LLC brings marginal performance gains in most applications. In other words, increasing the LLC size does not scale neither in performance nor energy consumption. We examine how LLC misses are managed in typical GPUs, and we find that in most cases the way LLC misses are managed are precisely the main performance limiter. This paper proposes a novel approach that addresses this shortcoming by leveraging a tiny additional Fetch and Replacement Cache-like structure (FRC) that stores control and coherence information of the incoming blocks until they are fetched from main memory. Then, the fetched blocks are swapped with the victim blocks (i.e., selected to be replaced) in the LLC, and the eviction of such victim blocks is performed from the FRC. This approach improves performance due to three main reasons: i) the lifetime of blocks being replaced is enlarged, ii) the main memory path is unclogged on long bursts of LLC misses, and iii) the average LLC miss latency is reduced. The proposal improves the LLC hit ratio, memory-level parallelism, and reduces the miss latency compared to much larger conventional caches. Moreover, this is achieved with reduced energy consumption and with much less area requirements. Experimental results show that the proposed FRC cache scales in performance with the number of GPU compute units and the LLC size, since, depending on the FRC size, performance improves ranging from 30 to 67 percent for a modern baseline GPU card, and from 32 to 118 percent for a larger GPU. In addition, energy consumption is reduced on average from 49 to 57 percent for the larger GPU. These benefits come with a small area increase (by 7.3 percent) over the LLC baseline., This work has been supported by the Spanish Ministerio de Ciencia, Innovacion y Universidades and the European ERDF under Grants T-PARCCA (RTI2018-098156-B-C51), and TIN2016-76635-C2-1-R (AEI/ERDF, EU), by the Universitat Politecnica de Valencia under Grant SP20190169, and by the gaZ: T58_17R research group (Aragon Gov. and European ESF).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

An Aging-Aware GPU Register File Design Based on Data Redundancy

  • Valero Bresó, Alejandro
  • Candel-Margaix, Francisco
  • Suárez-Gracia, Darío
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
[EN] Nowadays, GPUs sit at the forefront of high-performance computing thanks to their massive computational capabilities. Internally, thousands of functional units, architected to be fed by large register files, fuel such a performance. At deep nanometer technologies, the SRAM memory cells that implement GPU register files are very sensitive to the Negative Bias Temperature Instability (NBTI) effect. NBTI ages cell transistors by degrading their threshold voltage Vth over the lifetime of the GPU. This degradation, which manifests when a cell keeps the same logic value for a relatively long period of time, compromises the cell read stability and increases the transistor switching delay, which can lead to wrong read values and eventually exceed the processor cycle time, respectively, so resulting in faulty operation. Thiswork proposes architectural mechanisms leveraging the redundancy of the data stored in GPU register files to attack NBTI aging. The proposed mechanisms are based on data compression, power gating, and register address rotation techniques. All these mechanismsworking together balance the distribution of logic values stored in the cells along the execution time, reducing both the overall Vth degradation and the increase in the transistor switching delays. Experimental results show that a conventional GPU register file suffers the worst case for NBTI, since a significant fraction of the cells maintain the same logic value during the entire application execution (i.e., a 100 percent '0' and '1' duty cycle distributions). On average, the proposal reduces these distributions by 58 and 68 percent, respectively, which translates into Vth degradation savings by 54 and 62 percent, respectively., This work was supported by the Gobierno de Aragon and the European ESF (gaZ: T58_17R research group), and by the Ministerio de Economia y Competitividad (MINECO) and AEI/FEDER (EU) funds under Grants TIN2016-76635-C2-1-R and TIN2015-66972-C5-1-R.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publication Artículo científico (JournalArticle)

An aging-aware GPU register file design based on data redundancy

  • Valero Bresó, Alejandro
  • Candel Margaix, Francisco
  • Suárez Gracia, Darío
  • Petit Martí, Salvador Vicente
  • Sahuquillo Borrás, Julio
Nowadays, GPUs sit at the forefront of high-performance computing thanks to their massive computational capabilities. Internally, thousands of functional units, architected to be fed by large register files, fuel such a performance. At deep nanometer technologies, the SRAM memory cells that implement GPU register files are very sensitive to the Negative Bias Temperature Instability (NBTI) effect. NBTI ages cell transistors by degrading their threshold voltage Vth over the lifetime of the GPU. This degradation, which manifests when a cell keeps the same logic value for a relatively long period of time, compromises the cell read stability and increases the transistor switching delay, which can lead to wrong read values and eventually exceed the processor cycle time, respectively, so resulting in faulty operation. This work proposes architectural mechanisms leveraging the redundancy of the data stored in GPU register files to attack NBTI aging. The proposed mechanisms are based on data compression, power gating, and register address rotation techniques. All these mechanisms working together balance the distribution of logic values stored in the cells along the execution time, reducing both the overall Vth degradation and the increase in the transistor switching delays. Experimental results show that a conventional GPU register file suffers the worst case for NBTI, since a significant fraction of the cells maintain the same logic value during the entire application execution (i.e., a 100% ‘0’ and ‘1’ duty cycle distributions). On average, the proposal reduces these distributions by 58% and 68%, respectively, which translates into Vth degradation savings by 54% and 62%, respectively.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publication Artículo científico (JournalArticle)

Thermal-aware real-time scheduling using timed continuous Petri Nets

  • Desirena-López, G.
  • Ramírez-Treviño, A.
  • Briz Velasco, J.L.
  • Vázquez, C.R.
  • Gómez-Gutiérrez, D.
We present a thermal-aware, hard real-time (HRT) global scheduler for a multiprocessor system designed upon three novel techinques. First, we present a modeling methodology based on Timed Continuous Petri nets (TCPN) that yields a complete state variable model, including job arrivals, CPU usage, power, and thermal behavior. The model is accurate and avoids the calibration stage of RC thermal models. Second, based on this model, a linear programming problem (LPP) determines the existence of a feasible HRT thermal-aware schedule. Last, a sliding-mode controller and an online discretization algorithm implement the global HRT scheduler, which is capable of managing thermal constraints, context switching, migrations, and disturbances.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publication Artículo científico (JournalArticle)

A fault-tolerant last level cache for CMPs operating at ultra-low voltage

  • Ferrerón, A.
  • Alastruey-Benedé, J.
  • Suárez Gracia, D.
  • Monreal Arnal, T.
  • Ibáñez Marín, P.
  • Viñals Yúfera, V.
Voltage scaling to values near the threshold voltage is a promising technique to hold off the many-core power wall. However, as voltage decreases, some SRAM cells are unable to operate reliably and show a behavior consistent with a hard fault. Block disabling is a micro-architectural technique that allows low-voltage operation by deactivating faulty cache entries, at the expense of reducing the effective cache capacity. In the case of the last-level cache, this capacity reduction leads to an increase in off-chip memory accesses, diminishing the overall energy benefit of reducing the voltage supply. In this work, we exploit the reuse locality and the intrinsic redundancy of multi-level inclusive hierarchies to enhance the performance of block disabling with negligible cost. The proposed fault-aware last-level cache management policy maps critical blocks, those not present in private caches and with a higher probability of being reused, to active cache entries. Our evaluation shows that this fault-aware management results in up to 37.3% and 54.2% fewer misses per kilo instruction (MPKI) than block disabling for multiprogrammed and parallel workloads, respectively. This translates to performance enhancements of up to 13% and 34.6% for multiprogrammed and parallel workloads, respectively.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publication Artículo científico (JournalArticle)

Memory hierarchy characterization of SPEC CPU2006 and SPEC CPU2017 on the Intel Xeon Skylake-SP

  • Navarro-Torres, Agustín
  • Alastruey-Benedé, Jesús
  • Ibáñez-Marín, Pablo
  • Viñals-Yúfera, Víctor
SPEC CPU is one of the most common benchmark suites used in computer architecture research. CPU2017 has recently been released to replace CPU2006. In this paper we present a detailed evaluation of the memory hierarchy performance for both the CPU2006 and single-threaded CPU2017 benchmarks. The experiments were executed on an Intel Xeon Skylake-SP, which is the first Intel processor to implement a mostly non-inclusive last-level cache (LLC). We present a classification of the benchmarks according to their memory pressure and analyze the performance impact of different LLC sizes. We also test all the hardware prefetchers showing they improve performance in most of the benchmarks. After comprehensive experimentation, we can highlight the following conclusions: i) almost half of SPEC CPU benchmarks have very low miss ratios in the second and third level caches, even with small LLC sizes and without hardware prefetching, ii) overall, the SPEC CPU2017 benchmarks demand even less memory hierarchy resources than the SPEC CPU2006 ones, iii) hardware prefetching is very effective in reducing LLC misses for most benchmarks, even with the smallest LLC size, and iv) from the memory hierarchy standpoint the methodologies commonly used to select benchmarks or simulation points do not guarantee representative workloads.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publication Artículo científico (JournalArticle)

ReD: A reuse detector for content selection in exclusive shared last-level caches

  • Díaz, J.
  • Monreal, T.
  • Ibáñez, P.
  • Llabería, J.M.
  • Viñals, V.
The reference stream reaching a chip multiprocessor Shared Last-Level Cache (SLLC) shows poor temporal locality, making conventional cache management policies inefficient. Few proposals address this problem for exclusive caches. In this paper, we propose the Reuse Detector (ReD), a new content selection mechanism for exclusive hierarchies that leverages reuse locality at the SLLC, a property that states that blocks referenced more than once are more likely to be accessed in the near future. Being placed between each L2 private cache and the SLLC, ReD prevents the insertion of blocks without reuse into the SLLC. It is designed to overcome problems affecting similar recent mechanisms (low accuracy, reduced visibility window and detector thrashing). ReD improves performance over other state-of-the-art proposals (CHAR, Reuse Cache and EAF cache). Compared with the baseline system with no content selection, it reduces the SLLC miss rate (MPI) by 10.1% and increases harmonic IPC by 9.5%.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publication Artículo científico (JournalArticle)

Inference in supervised spectral classifiers for on-board hyperspectral imaging: An overview

  • Alcolea, Adrián
  • Paoletti, Mercedes E.
  • Haut, Juan M.
  • Resano, Javier
  • Plaza, Antonio
Machine learning techniques are widely used for pixel-wise classification of hyperspectral images. These methods can achieve high accuracy, but most of them are computationally intensive models. This poses a problem for their implementation in low-power and embedded systems intended for on-board processing, in which energy consumption and model size are as important as accuracy. With a focus on embedded anci on-board systems (in which only the inference step is performed after an off-line training process), in this paper we provide a comprehensive overview of the inference properties of the most relevant techniques for hyperspectral image classification. For this purpose, we compare the size of the trained models and the operations required during the inference step (which are directly related to the hardware and energy requirements). Our goal is to search for appropriate trade-offs between on-board implementation (such as model size anci energy consumption) anci classification accuracy.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publication Artículo científico (JournalArticle)

Reducing the WCET and analysis time of systems with simple lockable instruction caches

  • Pedro-Zapater, Alba
  • Segarra, Juan
  • Gran Tejero, Rubén
  • Viñals, Víctor
  • Rodríguez, Clemente
One of the key challenges in real-time systems is the analysis of the memory hierarchy. Many Worst-Case Execution Time (WCET) analysis methods supporting an instruction cache are based on iterative or convergence algorithms, which are rather slow. Our goal in this paper is to reduce the WCET analysis time on systems with a simple lockable instruction cache, focusing on the Lock-MS method. First, we propose an algorithm to obtain a structure-based representation of the Control Flow Graph (CFG). It organizes the whole WCET problem as nested subproblems, which takes advantage of common branch-and-bound algorithms of Integer Linear Programming (ILP) solvers. Second, we add support for multiple locking points per task, each one with specific cache contents, instead of a given locked content for the whole task execution. Locking points are set heuristically before outer loops. Such simple heuristics adds no complexity, and reduces the WCET by taking profit of the temporal reuse found in loops. Since loops can be processed as isolated regions, the optimal contents to lock into cache for each region can be obtained, and the WCET analysis time is further reduced. With these two improvements, our WCET analysis is around 10 times faster than other approaches. Also, our results show that the WCET is reduced, and the hit ratio achieved for the lockable instruction cache is similar to that of a real execution with an LRU instruction cache. Finally, we analyze the WCET sensitivity to compiler optimization, showing for each benchmark the right choices and pointing out that O0 is always the worst option.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publication Artículo científico (JournalArticle)

Energy-efficient thermal-aware multiprocessor scheduling for real-time tasks using TCPNs

  • Rubio-Anguiano, Laura
  • Desirena-López, Gaddiel
  • Ramírez-Treviño, Antonio
  • Briz Velasco, José Luis
We present an energy-effcient thermal-aware real-time global scheduler for a set of hard real-time (HRT) tasks running on a multiprocessor system. This global scheduler fulfills the thermal and temporal constraints by handling two independent variables, the task allocation time and the selection of clock frequency. To achieve its goal, the proposed scheduler is split into two stages. An off-line stage, based on a deadline partitioning scheme, computes the cycles that the HRT tasks must run per deadline interval at the minimum clock frequency to save energy while honoring the temporal and thermal constraints, and computes the maximum frequency at which the system can run below the maximum temperature. Then, an on-line, event-driven stage performs global task allocation applying a Fixed-Priority Zero-Laxity policy, reducing the overhead of quantum-based or interval-based global schedulers. The on-line stage embodies an adaptive scheduler that accepts or rejects soft RT aperiodic tasks throttling CPU frequency to the upper lowest available one to minimize power consumption while meeting time and thermal constraints. This approach leverages the best of two worlds: the off-line stage computes an ideal discrete HRT multiprocessor schedule, while the on-line stage manage soft real-time aperiodic tasks with minimum power consumption and maximum CPU utilization.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publication Artículo científico (JournalArticle)

Misterio de los Letrabytes Fugados: Evaluación Inclusiva de la Conciencia Fonológica con Gamificación Tangible

  • Oliva, Ana
  • Ayuso, Natalia
  • Coma, Teresa
  • Torres, Enrique
The development of phonological awareness, i.e. the metacognitive skill that allows thinking and manipulating the language structure, from an early age is mandatory for the acquisition of literacy skills. Multisensory stimulation, curiosity and interest are fundamental for learning. In doing so, Communication Technologies have a key role because a learning-oriented use can encourage current digital natives' interest and thinking. This paper presents a game with tangible technology designed to both intervention and diagnostic, formative and summative evaluation. The user experience was carried out during two weeks through a multiple case study with four kindergarten children. A multimedia adventure game was presented in which they had to solve phonological awareness challenges through the manipulation of tangible technology to achieve the final goal. The observation was recorded in the designed evaluation model, allowing the analysis of pre- and post-intervention data after the triangulation of data for subsequent qualitative analysis. The results show an improvement in phonological awareness, as well as the gamified experience marks. These promising outcomes support the use of such methodologies to a learner-centered education model.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publication Artículo científico (JournalArticle)

GPU-friendly neural networks for remote sensing scene classification

  • Haut, Juan M.
  • Alcolea, Adrian
  • Paoletti, Mercedes E.
  • Plaza, Javier
  • Resano, Javier
  • Plaza, Antonio
Convolutional neural networks (CNNs) have proven to be very efficient for the analysis of remote sensing (RS) images. Due to the inherent complexity of extracting features from these images, along with the increasing amount of data to be processed (and the diversity of applications), there is a clear tendency to develop and employ increasingly deep and complex CNNs. In this regard, graphics processing units (GPUs) are frequently used to optimize their execution, both for the training and inference stages, optimizing the performance of neural models through their many-core architecture. Hence, the efficient use of the GPU resources should be at the core of optimizations. This letter analyzes the possibilities of using a new family of CNNs, denoted as TResNets, to provide an efficient solution to the RS scene classification problem. Moreover, the considered models have been combined with mixed precision to enhance their training performance. Our experimental results, conducted over three publicly available RS data sets, show that the proposed networks achieve better accuracy and more efficient use of GPU resources than other state-of-the-art networks. Source code is available at https://github.com/mhaut/GPUfriendlyRS.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publication Artículo científico (JournalArticle)

Ideal and predictable hit ratio for matrix transposition in data caches

  • Pedro-Zapater, Alba
  • Rodríguez, Clemente
  • Segarra, Juan
  • Gran Tejero, Rubén
  • Viñals-Yúfera, Víctor
Matrix transposition is a fundamental operation, but it may present a very low and hardly predictable data cache hit ratio for large matrices. Safe (worst-case) hit ratio predictability is required in real-time systems. In this paper, we obtain the relations among the cache parameters that guarantee the ideal (predictable) data hit ratio assuming a Least-Recently-Used (LRU) data cache. Considering our analytical assessments, we compare a tiling matrix transposition to a cache oblivious algorithm, modified with phantom padding to improve its data hit ratio. Our results show that, with an adequate tile size, the tiling version results in an equal or better data hit ratio. We also analyze the energy consumption and execution time of matrix transposition on real hardware with pseudo-LRU (PLRU) caches. Our analytical hit/miss assessment enables the usage of a data cache for matrix transposition in real-time systems, since the number of misses in the worst case is bound. In general and high-performance computation, our analysis enables us to restrict the cache resources devoted to matrix transposition with no negative impact, in order to reduce both the energy consumption and the pollution to other computations.