Pasar al contenido principal

MEJORA DEL PROCESADOR, SUBSISTEMA DE MEMORIA, ACELERADORES Y REDES

PID2021-123627OB-C51

•
Nombre agencia financiadora Agencia Estatal de Investigación
Acrónimo agencia financiadora AEI
Programa Programa Estatal para Impulsar la Investigación Científico-Técnica y su Transferencia
Subprograma Subprograma Estatal de Generación de Conocimiento
Convocatoria Proyectos de I+D+i (Generación de Conocimiento y Retos Investigación)
Año convocatoria 2021
Plan Estatal 2021-2023 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario UNIVERSITAT POLITÈCNICA DE VALÈNCIA

Resultados relacionados

Resultados totales (Incluyendo duplicados): 49
Encontrada(s) 1 página(s)

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

SpinQ: Compilation strategies for scalable spin-qubit architectures

  • Paraskevopoulos, Nikiforos
  • Sebastiano, Fabio
  • Feld, Sebastian
  • GARCIA ALMUDEVER, CARMEN|||0000-0002-3800-2357
[EN] Despite Noisy Intermediate-Scale Quantum devices being severely constrained, hardware- and algorithm-aware quantum circuit mapping techniques have been developed to enable successful algorithm executions. Not so much attention has been paid to mapping and compilation implementations for spin-qubit quantum processors due to the scarce availability of experimental devices and their small sizes. However, based on their high scalability potential and their rapid progress it is timely to start exploring solutions on such devices. In this work, we discuss the unique mapping challenges of a scalable crossbar architecture with shared control and introduce SpinQ, the first native compilation framework for scalable spin-qubit architectures. At the core of SpinQ is the Integrated Strategy that addresses the unique operational constraints of the crossbar while considering compilation scalability and obtaining a O(n) computational complexity. To evaluate the performance of SpinQ on this novel architecture, we compiled a broad set of well-defined quantum circuits and performed an in-depth analysis based on multiple metrics such as gate overhead, depth overhead, and estimated success probability, which in turn allowed us to create unique mapping and architectural insights. Finally, we propose novel mapping techniques that could increase algorithm success rates on this architecture and potentially inspire further research on quantum circuit mapping for other scalable spin-qubit architectures., This work is part of the research program OTP with project number 16278, which is (partly) financed by the Netherlands Organisation for Scientific Research (NWO). This work has also been partially supported by the Spanish Ministerio de Ciencia e Innovación, European ERDF under grant PID2021-123627OB-C51 (CGA). We thank Menno Veldhorst and Hans van Someren for their fruitful discussions.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Diseño y Exploración de Aceleradores Hardware en la Plataforma SELENE

  • Andreu-Cerezo, Pablo|||0000-0003-2070-9271
  • López Rodríguez, Pedro Juan|||0000-0003-4544-955X
  • Flich Cardo, José
  • Hernández Luz, Carles|||0000-0003-1192-6223
  • Picornell, Tomás
  • Kostalampros, Vatistas
  • Moreto, Miquel
[ES] El diseño de aceleradores hardware esta adquiriendo una gran importancia debido al aumento de aplicaciones que usan redes neuronales y la aproximacion al l¿mite de la ley de Moore. En este contexto, la exploracion de las opciones de diseño de estos aceleradores y su impacto en el rendimiento necesita de plataformas donde integrar dichos aceleradores de manera agil. Con esa finalidad este trabajo presenta la plataforma de SELENE, una plataforma abierta donde es sencillo integrar aceleradores diseñados en Xilinx HLS o con interfaz estandar AXI para su testeo en FPGA., El presente trabajo ha sido parcialmente financiado mediante el proyecto IPMAN (Tecnicas Innovadoras para Infraestructuras, Aplicaciones y Servicios en Centros de Datos y Sistemas Altamente Distribuidos) con referencia PID2021-123627OB-C51 y por la Generalitat Valenciana mediante el proyecto Subvenciones para la contratacion de personal investigador predoctoral con codigo CIACIF/2021/412.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

SYNPA: SMT Performance Analysis and Allocation of Threads to Cores in ARM Processors

  • Navarro-Edo, Marta|||0000-0002-4334-460X
  • Feliu-Pérez, Josué|||0000-0003-3017-4266
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
[EN] Simultaneous multithreading processors improve throughput over single-threaded processors thanks to sharing internal core resources among instructions from distinct threads. However, resource sharing introduces inter-thread interference within the core, which has a negative impact on individual application performance and can significantly increase the turnaround time of multi-program workloads. The severity of the interference effects depends on the competing co-runners sharing the core. Thus, it can be mitigated by applying a thread-to-core allocation policy that smartly selects applications to be run in the same core to minimize their interference. This paper presents SYNPA, a simple approach that dynamically allocates threads to cores in an SMT processor based on their run-time dynamic behavior. The approach uses a regression model to select synergistic pairs to mitigate intracore interference. The main novelty of SYNPA is that it uses just three variables collected from the performance counters available in current ARM processors at the dispatch stage. Experimental results show that SYNPA outperforms the default Linux scheduler by around 36%, on average, in terms of turnaround time in 8-application workloads combining frontendbound and backend-bound benchmarks. Index Terms¿Simultaneous Multithreading, Scheduling, HighPerformance Computing, HPC applications, Thread-to-Core allocation, synergistic applications, intra-core interference., This work has been supported by the Spanish Ministerio de Ciencia e Innovacion and European ERDF under grant PID2021-123627OB-C51, TED2021-130233B-C32 and Generalitat Valenciana under grant AICO/2021/266. Marta Navarro is supported by Subvenciones para la contratacion de personal investigador predoctoral by CIACIF/2021/413 . Josue Feliu is supported by the RYC2021-030862-I contract, funded by the MCIN/AEI /10.13039/501100011033 and the European Union NextGenerationEU/PRTR.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Tesis de maestría (MasterThesis)

Characterization of Performance and Energy Efficiency in High-Performance Processors, Caracterización de prestaciones y eficiencia energética en procesadores de altas prestaciones, Caracterització de prestacions i eficiència energètica en processadors d&apos, altes prestacions

  • Calero Quintana, Ibai
[ES] La creciente importancia de la eficiencia energética ha impulsado la necesidad de desarrollar estrategias de planificación que reduzcan el consumo de energía tratando de mantener las prestaciones en todo tipo de dispositivos, incluidos servidores y supercomputadores. Las aplicaciones secuenciales y paralelas representan dos enfoques distintos en la ejecución de tareas. Las aplicaciones secuenciales disponen de un único hilo de ejecución, mientras que las aplicaciones paralelas dividen las tareas en múltiples hilos que operan simultáneamente, permitiendo un procesamiento concurrente. Las aplicaciones secuenciales presentan desafíos en la gestión de las interferencias en los recursos compartidos del procesador. Es necesario identificar qué recursos estresan y cómo sus prestaciones se ven afectadas por estas interferencias. Por el contrario, las aplicaciones paralelas plantean retos adicionales, como determinar el número óptimo de hilos a asignar para maximizar sus prestaciones sin causar grandes incrementos en el consumo. El presente trabajo busca caracterizar comportamientos típicos de las aplicaciones y analizar las correlaciones entre diversas métricas de rendimiento y consumo energético, con el objetivo de facilitar el diseño de nuevas propuestas de planificación. Del análisis se desprende que las aplicaciones secuenciales se pueden clasificar según cómo sus prestaciones se ven afectadas por interferencias en los recursos compartidos. Las aplicaciones que sufren menos las interferencias en los recursos compartidos ofrecen mejores prestaciones a costa de un mayor consumo, mientras que aquellas con más interferencias presentan peores prestaciones pero un consumo menor. En cuanto a las aplicaciones paralelas, se identifican varios comportamientos en función de cómo sus prestaciones y consumo energético se ven afectados al aumentar el número de hilos. Se observa que algunas aplicaciones se benefician mucho de un número elevado de hilos, mejorando sus prestaciones mucho más que su consumo, mientras que otras no son capaces de aprovechar un número elevado de hilos, resultando en un escaso incremento en las prestaciones pero elevados consumos debido a un elevado número de hilos., [EN] The growing importance of energy efficiency has driven the need to develop scheduling approaches that reduce energy consumption while trying to maintain performance across all types of devices, including servers and supercomputers. Sequential and parallel applications represent two distinct approaches to task execution. Sequential applications have a single execution thread, whereas parallel applications divide tasks into multiple threads that operate simultaneously, allowing concurrent processing. Sequential applications present challenges in managing interference in the processor's shared resources. It is necessary to identify which resources they stress and how their performance is affected by this interference. Conversely, parallel applications pose additional challenges, such as determining the optimal number of threads to assign to maximize their performance without causing significant increases in consumption. This work aims to characterize typical application behaviors and analyze the correlations between various performance and energy consumption metrics, with the goal of facilitating the design of new scheduling proposals. Results reveal that sequential applications can be classified based on how their performance is affected by interference in shared resources. Applications less affected by interferences in shared resources offer better performance at the cost of higher consumption, whereas those more influenced by interferences exhibit poorer performance but lower consumption. Regarding parallel applications, several behaviors are identified based on how their performance and energy consumption are affected by increasing the number of threads. It is observed that some applications benefit greatly from a high number of threads, increasing their performance significantly more than their consumption, while others are unable to take advantage of a high number of threads, resulting in minimal performance gains but high consumption due to the large number of active threads.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Dynamic Allocation of Processor Cores to Graph Applications on Commodity Servers

  • Pons-Escat, Lucía|||0000-0002-4582-7744
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • Jones, Timothy M.
[EN] Graph processing is increasingly adopted to solve problems that span many application domains, including scientific computing, social networks, and big-data analytics. These applications present particular features (huge working sets and irregular scalability) that make the default Linux scheduler, which adopts a time-sharing policy to provide a fair scheduler, perform poorly when co-locating multiple graph applications in the same processor. This work focuses on maximizing processor utilization, which is a major concern of current data centers. To this end, we propose AFAIR, a flexible scheduling policy that allocates multiple graph applications on the same processor and assigns a fraction of the cores exclusively to each application instead of sharing them. Moreover, AFAIR dynamically adds/removes cores to the running applications, adapting the number of threads used for parallel execution to balance memory load. This allows AFAIR to achieve almost perfect fairness, on average 95%., This work has been supported by the Spanish Ministerio de Universidades under the grant FPU18/01948 and 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
Publicación Contribución de congreso (ConferenceOutput)

Mejora del Benchmarking de aplicaciones de latencia crítica para una evaluación más realista

  • Li, Zhilin
  • Pons-Escat, Lucía|||0000-0002-4582-7744
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • Pons Terol, Julio|||0000-0002-5654-6753
[ES] Los sistemas en la nube han crecido rápidamente, trasladando tareas computacionales desde los dispositivos usuarios a los servidores de la nube donde los clientes envían sus peticiones. Las aplicaciones de latencia crítica, influenciadas por los tiempos de respuesta más elevados (percentil 95+) de las peticiones, han cobrado especial interés, pero su gran diversidad dificulta una evaluación precisa. Este trabajo compara distintas suites de prueba y concluye que TailBench es la más adecuada para evaluar estas aplicaciones, ya que ofrece una amplia diversidad de comportamientos e implementa una interfaz común que facilita la evaluación experimental. Sin embargo, presenta limitaciones para escenarios multi-servidor. Para abordar este problema, en este trabajo se propone TailBench++, una versión mejorada de Tailbench que permite configurar de forma flexible el número de clientes y servidores, así como la carga de trabajo. Estas mejoras facilitan la evaluación de entornos más realistas con múltiples servidores que sirven peticiones de clientes que varían su comportamiento de forma dinámica, ampliando el alcance de los estu- dios en la nube., Este trabajo ha sido financiado por el Ministerio de Ciencia e Innovacion junto al FEDER europeo a traves de los proyectos PID2021-123627OB-C51 y TED2021-130233B-C32.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

ArtA: automating Design Space Exploration of spin-qubit architectures

  • Paraskevopoulos, Nikiforos
  • Hamel, David
  • Sarkar, Aritra
  • Feld, Sebastian
  • GARCIA ALMUDEVER, CARMEN|||0000-0002-3800-2357
[EN] In the fast-paced field of quantum computing, identifying the architectural characteristics that will enable quantum processors to achieve high performance across a diverse range of quantum algorithms continues to pose a significant challenge. Given the extensive and costly nature of experimentally testing different designs, this paper introduces the first Design Space Exploration (DSE) for quantum-dot spin-qubit architectures. Utilizing the upgraded SpinQ compilation framework, this study explores a substantial design space comprising 29,312 spin-qubit-based architectures and applies an innovative optimization tool, ArtA (Artificial Architect), to speed up the design space traversal. ArtA can leverage 17 optimization configurations, significantly reducing exploration times by up to 99.1% compared to a traditional brute force approach while maintaining the same result quality. After a comprehensive evaluation of best-matching optimization configurations per quantum circuit, ArtA suggests specific as well as universal architectural features that provide optimal performance across the examined circuits. Our work demonstrates that combining DSE methodologies with optimization algorithms can be effectively used to generate meaningful design insights for quantum processor development., This work is part of the research program OTP with project number 16278, which is (partly) financed by the Netherlands Organisation for Scientific Research (NWO). CGA acknowledges funding from the Spanish Ministry of Science, Innovation and Universities through the Beatriz Galindo program 2020 (BG20-00023) and from the European ERDF under grant PID2021-123627OB-C51.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Characterizing Power and Performance Interference Scalability in the 28-core 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] Nowadays, energy efficiency is a major concern in any type of processor-based device, ranging from processor servers to supercomputers, including mobile battery-fed devices. In recent years, systems based on the ARM architecture, traditionally better suited for mobile and embedded systems, have increased their market share in segments commonly occupied by x86 processors. This growth is due, to some extent, to the excellent energy efficiency shown by high-performance ARM processors. Designing software and hardware energy-efficient systems requires a sound knowledge of the relationship among three main axes: component activity, power, and inter-application interference at the shared resources. This problem aggravates in many-core processors, which are ubiquitous in high-performance servers. This paper characterizes the aforementioned axes in a 28-core ARM ThunderX2 processor. Experimental results show that the performance of highly-scalable single-threaded applications is sustained regardless of the number of applications and interference introduced at the shared resources at the cost of increasing power. In contrast, low-scalable applications require much less power and experience a huge performance degradation as their number increases., 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
Publicación Contribución de congreso (ConferenceOutput)

Compilation Techniques for Spin Qubits in a Shuttling Bus Architecture

  • Escofet, Pau
  • Semenov, Andrii
  • Murphy, Niall
  • Blokhina, Elena
  • Abadal, Sergi
  • Alarcon, Eduard
  • GARCIA ALMUDEVER, CARMEN|||0000-0002-3800-2357
[EN] In this work, we explore and propose several quantum circuit mapping strategies to optimize qubit shuttling in scalable quantum computing architectures based on silicon spin qubits. Our goal is to minimize phase errors introduced during shuttling operations while reducing the overall execution time of quantum circuits. We propose and evaluate five mapping algorithms using benchmarks from quantum algorithms. The Swap Return strategy emerged as the most robust solution, offering a superior balance between execution time and error minimization by considering future qubit interactions. Additionally, we assess the importance of initial qubit placement, demonstrating that an informed placement strategy can significantly enhance the performance of dynamic mapping approaches., This work was supported by the European Commission (QUADRATURE: HORIZON-EIC-2022-PATHFINDEROPEN-01-101099697, WINC: HORIZON-ERC-2021-101042080), the Spanish Ministry of Science, Innovation and Universities (Beatriz Galindo program 2020, BG20-00023) and European ERDF (PID2021-123627OB-C5, QCOMM-CAT-Planes Complementarios), funding from MICIN and NextGenerationEU (PRTR-C17.I1), Generalitat de Catalunya, the ICREA Academia Award 2024, Science Foundation Ireland under Grant 21/RP-2TF/10019, and the predoctoral FPI-UPC grant supported by UPC and Banco Santander.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Thread-to-Core Allocation in ARM Processors Buildings Synergistic Pairs

  • Navarro-Edo, Marta|||0000-0002-4334-460X
  • Feliu-Pérez, Josué|||0000-0003-3017-4266
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
[EN] Simultaneous multithreading (SMT) processors can present significant throughput improvements over singlethreaded (ST) processors thanks to sharing internal core resources among instructions executing from multiple threads. However, resource sharing introduces inter-thread interference within the core, which negatively impacts individual application performance and can significantly increase the turnaround time of multi-program workloads. The severity of the intra-core interference on performance depends on the applications corunning in the same core. A thread scheduler can help reduce this effect by smartly selecting the pairs of applications that should run on each SMT core. This paper presents SYNPA, a simple approach that dynamically allocates threads to SMT cores based on their run-time dynamic behavior. SYNPA uses a regression model to select synergistic pairs to mitigate intra-core interference. Results show that SYNPA outperforms the default Linux scheduler by around 35%, on average, in terms of turnaround time when running 8- application workloads combining frontend-bound and backendbound applications., This work has been supported by the Spanish Ministerio de Ciencia e Innovacion and European ERDF under grant PID2021-123627OB-C51, TED2021-130233B-C32, and Generalitat Valenciana under grant AICO/2021/266. Marta Navarro is supported by Subvenciones para la contratacion de personal investigador predoctoral by CIACIF/2021/413 . Josue Feliu is supported by the RYC2021-030862-I contract, funded by the MCIN/AEI /10.13039/501100011033 and the European Union NextGenerationEU/PRTR.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación 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
Publicación Artículo científico (JournalArticle)

BeSnake: A Routing Algorithm for Scalable Spin-Qubit Architectures

  • Paraskevopoulos, Nikiforos
  • Feld, Sebastian
  • GARCIA ALMUDEVER, CARMEN|||0000-0002-3800-2357
[EN] As quantum computing devices increase in size with respect to the number of qubits, two-qubit interactions become more challenging, necessitating innovative and scalable qubit routing solutions. In this work, we introduce beSnake, a novel algorithm specifically designed to address the intricate qubit routing challenges in scalable spin-qubit architectures. Unlike traditional methods in superconducting architectures that solely rely on swap operations, beSnake also incorporates the shuttle operation to optimize the execution time and fidelity of quantum circuits and achieves fast computation times of the routing task itself. Employing a simple breadth-first search approach, beSnake effectively manages the restrictions created by diverse topologies and qubit positions acting as obstacles for up to 72% qubit density. It also has the option to adjust the level of optimization and to dynamically tackle parallelized routing tasks, all the while maintaining noise awareness. Our simulations demonstrate beSnake's advantage over an existing routing solution on random circuits and real quantum algorithms with up to 1000 qubits, showing an average improvement of up to 80% in gate overhead, 54% in depth overhead, and up to 8.33 times faster routing times., This work was supported in part by the Netherlands Organisation for Scientific Research through the Research Program OTP under Grant 16278. The work of Carmen G. Almudever was supported in part by the Spanish Ministry of Science, Innovation and Universities through the Beatriz Galindo Program 2020 under Grant BG20-00023 and in part by the European Regional Development Fund under Grant PID2021-123627OB-C51.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Scaling and Assigning Resources on ION Trap QCCD Architectures

  • Ovide-González, Anabel
  • GARCIA ALMUDEVER, CARMEN|||0000-0002-3800-2357
  • Cuomo, Daniele
[EN] Ion trap technologies have earned significant attention as potential candidates for quantum information processing due to their long decoherence times and precise manipulation of individual qubits, distinguishing them from other candidates in the field of quantum technologies. However, scalability remains a challenge, as introducing additional qubits into a trap increases noise and heating effects, consequently decreasing operational fidelity. Trapped-ion Quantum Charge-Coupled Device (QCCD) architectures have addressed this limitation by interconnecting multiple traps and employing ion shuttling mechanisms to transfer ions among traps. This new architectural design requires the development of novel compilation techniques for quantum algorithms, which efficiently allocate and route qubits, and schedule operations. The aim of a compiler is to minimize ion movements and, therefore, reduce the execution time of the circuit to achieve a higher fidelity. In this paper, we propose a novel approach for initial qubit placement, demonstrating enhancements of up to 50% compared to prior methods. Furthermore, we conduct a scalability analysis on two distinct QCCD topologies: a 1D-linear array and a ring structure. Additionally, we evaluate the impact of the excess capacity ¿ i.e. the number of free spaces within a trap ¿ on the algorithm performance., The authors acknowledge financial support from the European Union s Horizon Europe research and innovation program through the project Quantum Internet Alliance under grant agreement No. 101102140. CGA also acknowledges support from the Spanish Ministry of Science, Innovation and Universities through the Beatriz Galindo program 2020 (BG20-00023) and the European ERDF under grant PID2021- 123627OB-C51.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Extensión de un simulador de redes de interconexión y modelado de políticas de mapeo en redes Dragonfly

  • Baviera-Guaita, Alejandro
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
[ES] La red de interconexion es uno de los componentes mas cr¿ticos de los supercomputadores. La eleccion de la topolog¿a es cr¿tica para cubrir las necesidades de prestaciones y consumo de estos sistemas. Se suelen utilizan topolog¿as novedosas con conmutadores de alto grado como la Dragonfly. No obstante, esta topolog¿a no esta exenta de problemas de contencion en la red entre aplicaciones que comparten el sistema, as¿ como entre los mensajes de la misma aplicacion. Los trabajos que se ejecutan concurrentemente se ven obligados a compartir recursos generando interferencias entre ellos en las comunicaciones a traves de la red. La interferencia sufrida por las aplicaciones paralelas es altamente dependiente de las pol¿ticas de mapeo de aplicaciones a nodos de procesamiento en el sistema y de la configuracion de la topolog¿a utilizada. Por ello en este art¿culo analizamos en la topolog¿as Dragonfly la interferencia sufrida por aplicaciones reales del dominio HPC considerando varias estrategias de mapeo de estas aplicaciones en los nodos de procesamiento y varias configuraciones de la topolog¿a Dragonfly. Ademas, este art¿¿culo presenta las extensiones realizadas al simulador Sauron para poder realizar un analisis exhaustivo de interferencia en la red de interconexion de altas prestaciones en aplicaciones paralelas., Este trabajo ha sido parcialmente financiado por por el Ministerio de Ciencia e Innovacion junto al FEDER europeo a traves de los proyectos PID2021- 123627OB-C51 and TED2021-130233B-C32.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Revisiting the Mapping of Quantum Circuits: Entering the Multi-Core Era

  • Escofet, Pau
  • Bandic, Medina
  • Prielinger, Luise
  • Van Someren, Hans
  • Feld, Sebastian
  • Alarcon, Eduard
  • Abadal, Sergi
  • Ovide-González, Anabel
  • GARCIA ALMUDEVER, CARMEN|||0000-0002-3800-2357
[EN] Quantum computing represents a paradigm shift in computation, offering the potential to solve complex problems intractable for classical computers. Although current quantum processors already consist of a few hundred qubits, their scalability remains a significant challenge. Modular quantum computing architectures have emerged as a promising approach to scale up quantum computing systems. This article delves into the critical aspects of distributed multi-core quantum computing, focusing on quantum circuit mapping, a fundamental task to successfully execute quantum algorithms across cores while minimizing inter-core communications. We derive the theoretical bounds on the number of non-local communications needed for random quantum circuits and introduce the Hungarian Qubit Assignment (HQA) algorithm, a multi-core mapping algorithm designed to optimize qubit assignments to cores with the aim of reducing inter-core communications. Our exhaustive evaluation of HQA against state-of-the-art circuit mapping algorithms for modular architectures reveals a 4.9x and 1.6x improvement in terms of execution time and non-local communications, respectively, compared to the best-performing algorithm. HQA emerges as a very promising scalable approach for mapping quantum circuits into multi-core architectures, positioning it as a valuable tool for harnessing the potential of quantum computing at scale., The authors gratefully acknowledge funding from the European Commission through HORIZON-EIC-2022-PATHFINDEROPEN-01-101099697 (QUADRATURE) and grant HORIZON-ERC-2021-101042080 (WINC) . This work has been partially supported by the Spanish Ministerio de Ciencia e Innovacion and European ERDF under grant PID2021-123627OB-C5.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Synthesizing an optimal spin-qubit shuttling-bus architecture for the surface code

  • Escofet, Pau
  • Alarcon, Eduard
  • Abadal, S
  • Semenov, Andrii
  • Murphy, Niall
  • Blokhina, Elena
  • GARCIA ALMUDEVER, CARMEN|||0000-0002-3800-2357
[EN] As quantum computers scale toward millions of physical qubits, it becomes essential to robustly encode individual logical qubits to ensure fault tolerance under realistic noise. A high-quality foundational encoding allows future compilation techniques and heuristics to build on optimal or near-optimal layouts, improving scalability and error resilience. In this work, we synthesize a one-dimensional shuttling-bus architecture for the rotated surface code, leveraging coherent spin-qubit shuttling, following a methodology we name quantum reverse mapping. We formulate a mixed-integer optimization model that yields optimal solutions with relatively low execution time for small code distances, and propose a scalable heuristic that matches optimal results while maintaining linear computational complexity. We evaluate the synthesized architecture using architectural metrics, such as shuttling distance and cycle time, and full quantum simulations under realistic noise models, showing that the proposed design can sustain logical error rates as low as 2 & times; 10-10 per round at code distance 21, showcasing its feasibility for scalable quantum error correction in spin-based quantum processors., Authors acknowledge funding from the EC through Grant No. HORIZON-EIC-2022-PATHFINDEROPEN-01-101099697 (QUADRATURE) . E.A. acknowledges funding from the Spanish Ministry of Science, Innovation, and Universities and European ERDF (QCOMM-CAT-Planes Complementarios) , Generalitat de Catalunya, and ICREA Academia Grant 2024. S.A. acknowledges funding from the EC through Grant No. HORIZON-ERC-2021-101042080 (WINC) . E.B. acknowledges funding from Science Foundation Ireland under Grant No. 21/RP-2TF/10019. C.G.A. acknowledges funding from QuantERA EQUIP Grant No.PCI2022-133004, funded by Agencia Estatal de Investigacion, Ministerio de Ciencia e Innovacion, Gobierno de Espana, Project No. MCIN/AEI/10.13039/501100011033, from the European Union NextGenerationEU/PRTR, and from the Spanish Ministry of Science, Innovation, and Universities and European ERDF under Grants No. PID2021-123627OB-C51 and No. PID2024-158682OB-C31. P.E. acknowledges funding from the predoctoral FPI-UPC grant supported by UPC and Banco Santander.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

TailBench plus plus : flexible multi-client, multi-server benchmarking for latency-critical workloads

  • Li, Zhibin
  • Pons-Escat, Lucía|||0000-0002-4582-7744
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Pons Terol, Julio|||0000-0002-5654-6753
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
[EN] Cloud systems have rapidly expanded worldwide in the last decade, shifting computational tasks to cloud servers where clients submit their requests. Among cloud workloads, latency-critical applications -characterized by high-percentile response times - have gained special interest. These applications are present in modern services, representing an important fraction of cloud workloads. This work analyzes common cloud benchmarking suites and identifies TailBench as the most suitable to assess cloud performance with latency-critical workloads. Unfortunately, this suite presents key limitations, especially in multi-server scenarios or environments with variable client arrival patterns and fluctuating loads. To address these limitations, we propose TailBench++, an enhanced benchmark suite that extends TailBench to enable cloud evaluation studies to be performed in dynamic multi-client, multi-server environments. It allows reproducing experiments with varying client arrival times, dynamic query per second (QPS) fluctuations, and multiple servers handling requests. Case studies show that TailBench++ enables more realistic evaluations by capturing a wider range of real-world scenarios., This work has been supported by the Spanish Ministerio de Ciencia e Innovacion and European ERDF under grants PID2021-123627OB-C51 and PID2024-158682OB-C31, as well as by the Generalitat Valenciana under grant CIPROM/2024/007.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación 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
Publicación 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
Publicación Contribución de congreso (ConferenceOutput)

Caracterización de las prestaciones y la polución en las caches no inclusivas

  • Pons-Escat, Lucía|||0000-0002-4582-7744
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Pons Terol, Julio|||0000-0002-5654-6753
[ES] La jerarquía de cache ha estado en continua evolución en los procesadores generación tras generación. La última tendencia ha sido la redistribución del espacio de las estructuras de la jerarquía. Se ha reducido el área de la cache disponible por núcleo pero adoptando un esquema no inclusivo en la cache de último nivel. Como resultado, esta nueva organización disminuye el área total de cache disponible por núcleo. Por ello, la gestión del espacio de cache en esta organización se vuelve más crítica que en las caches inclusivas. Este artículo presenta un análisis detallado del impacto que tiene una cache de último nivel (L3) no inclusiva. Se presentan tres estudios de caracterización de las aplicaciones SPEC CPU2006 y SPEC CPU2017. Primero, se compara el reciente esquema no inclusivo con el esquema inclusivo. En segundo lugar, se analiza la relación entre el espacio de cache no inclusiva y las prestaciones -tiempo de ejecución- de las aplicaciones. Finalmente, se analiza la polución de la cache L3, donde se identifican y cuantifican tres principales fuentes de polución: useless prefetches, bad speculation, y poor locality. Los resultados experimentales muestran que las aplicaciones cuyas prestaciones son más sensitivas al espacio de cache de L3 experimentan un número de stalls_L2 y HPKI_L3 alto. Además, es posible identificar aquellas aplicaciones que introducen polución usando los contadores de prestaciones para así poder limitarles el espacio de cache., Este trabajo ha sido financiado por el Ministerio de Ciencia, Innovacion y Universidades mediante la beca FPU18/01948 y junto al FEDER europeo por el proyecto RTI2018-098156-B-C51.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Del mundo real al aula: Integrando la gestión de recursos de sistemas reales en la docencia de arquitectura del computador

  • Pons-Escat, Lucía|||0000-0002-4582-7744
  • Navarro-Edo, Marta|||0000-0002-4334-460X
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
[ES] La gestión de recursos es fundamental para optimizar las prestaciones y la equidad en los servidores actuales. Para desarrollar sistemas eficientes, es esencial una formación sólida en técnicas avanzadas de gestión de recursos. Sin embargo, las asignaturas avanzadas sobre arquitectura de computadores suelen concentrarse en los aspectos teóricos o prácticas relativamente sencillas, la mayoría basadas en simuladores, y omitir estos conceptos. Este artículo presenta la asignatura de máster "Gestión de Recursos en Sistemas Cloud y Evaluación de Prestaciones" (GRE) que aborda la arquitectura de computadores con un enfoque práctico, centrándose en la evaluación de prestaciones y la gestión de recursos de las máquinas comerciales. Los contenidos abarcan tanto sistemas de alto rendimiento (HPC) como sistemas cloud, donde una gestión ineficiente puede comprometer los acuerdos de nivel de servicio (SLA). Un aspecto fundamental es dotar a los alumnos de habilidades prácticas, por lo que se han diseñado prácticas en sistemas reales y empleando herramientas populares tanto en la industria como en la investigación. Esto asegura una formación aplicada y relevante, respaldada por altos niveles de satisfacción estudiantil., Este trabajo ha sido financiado por el Ministerio de Ciencia e Innovacion junto al FEDER europeo a traves de los proyectos PID2021-123627OB-C51 y TED2021-130233B-C32.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Predicting the Temporal Variability of Error Rates in Superconducting Quantum Processors: Mitigating Error Accuracy Loss in Queue-Based Quantum Computing Platforms [Focus: Quantum Software and its Engineering]

  • Rodríguez-Soriano, Laura|||0000-0002-2074-0466
  • García-Herrero, Francisco
  • GARCIA ALMUDEVER, CARMEN|||0000-0002-3800-2357
[EN] Multiple layers of the quantum system full-stack require data on temporal variations in error rates of physical elements. This article proposes using five machine learning and time-series models to predict error rate variability for two-qubit gates based on calibration data., This work is supported by the QuantERA grant EQUIP with Grants PCI2022-133004 and PCI2022-132922, funded by Agencia Estatal de Investigacion, Ministerio de Ciencia e Innovacion, Gobierno de Espana, MCIN/AEI /10.13039/501100011033, and by the European Union "NextGenerationEU/PRTR." CGA acknowledges funding from the Spanish Ministry of Science, In-novation, and Universities through the Beatriz Galindo program 2020 (BG20-00023), from the European ERDF under Grant PID2021-123627OB-C51, and from the Ministry for Digital Transformation and of Civil Service of the Spanish Government through the QUANTUM ENIA project call-Quantum Spain project, and by the European Union through the Recovery, Transformation and Resilience Plan - NextGenerationEU within the framework of the Digital Spain 2026 Agenda. We acknowledge the use of IBM Quantum services for this work. The views expressed are those of the authors and do not reflect the official policy or position of IBM or the IBM Quantum team.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Caracterización de prestaciones, interferencias y análisis de consumo energético en un procesador ARM Thunder X2

  • 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
[ES] En la actualidad, la eficiencia energética es de gran importancia en todo tipo de dispositivos, desde servidores hasta supercomputadores, pasando por ordenadores de escritorio. En este contexto, en los últimos años, los sistemas basados en la arquitectura ARM, tradicionalmente relegados a sistemas móviles y empotrados, han ganado una mayor cuota de mercado en nichos copados por procesadores basados en la arquitectura x86. Esta irrupción se debe a que los procesadores ARM de altas prestaciones ofrecen una mayor eficiencia energética que la competencia. El presente trabajo persigue caracterizar comportamientos t¿¿picos de las aplicaciones y analizar las correlaciones entre diversas métricas de prestaciones y consumo energético para servir de ayuda en el diseño de nuevas propuestas de planificación centradas en procesadores ARM. La principal conclusión del análisis de los resultados indica que las aplicaciones que sufren un menor slowdown, como consecuencia de la contención por los recursos compartidos del procesador, son también las que alcanzan una mayor eficiencia energética., Este trabajo ha sido financiado por el Ministerio de Ciencia e Innovaci¿on junto al FEDER europeo a traves de los proyectos TED2021-130233B-C32 y PID2021-123627OB-C51.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Energy efficient HPC network topologies with on/off links

  • Andújar-Muñoz, Francisco José
  • Sánchez García, José Luis
  • Alfaro Cortés, Francisco José
  • Coll, Salvador|||0000-0002-7795-7252
  • Alonso Díaz, Marina|||0000-0001-7079-5898
  • Martínez-Rubio, Juan-Miguel|||0000-0002-9748-0332
  • López Rodríguez, Pedro Juan|||0000-0003-4544-955X
[EN] Energy efficiency is a must in today HPC systems. To achieve this goal, a holistic design based on the use of power-aware components should be performed. One of the key components of an HPC system is the high-speed interconnect. In this paper, we compare and evaluate several design options for the interconnection network of an HPC system, including torus, fat-trees and dragonflies. State of the art low power modes are also used in the interconnection networks. The paper does not only consider energy efficiency at the interconnection network level but also at the system as a whole.The analysis is performed by using a simple yet realistic power model of the system. The model has been adjusted using actual power consumption values measured on a real system. Using this model, realistic multi-job trace-based workloads have been used, obtaining the execution time and energy consumed. The results are presented to ease choosing a system, depending on which parameter, performance or energy consumption, receives the most importance., This work has been supported by the Spanish Ministerio de Ciencia e Innovacion (MICINN, formerly MINECO) , and the European Commission (FEDER funds) under the projects PID2019- 105903RB-100 and PID2021-123627OB-C5, and by Junta de Comunidades de Castilla -La Mancha under the project SBPLY/21/180501/000248.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Route-Forcing: Scalable Quantum Circuit Mapping for Scalable Quantum Computing Architectures

  • Escofet, Pau
  • Gonzalvo, Alejandro
  • Alarcon, Eduard
  • Abadal, Sergi
  • GARCIA ALMUDEVER, CARMEN|||0000-0002-3800-2357
[EN] Quantum computers are expected to scale in size to close the gap that currently exists between quantum algorithms and quantum hardware. To this end, quantum compilation techniques must scale along with the hardware constraints, shifting the current paradigm of obtaining an optimal compilation to relying on heuristics that allow for a fast solution, even though the quality of such a solution may not be optimal. Significant concerns arise as the execution time of current mapping techniques experiences a notable increase when applied to quantum computers with a high number of qubits. In this work, we present Route-Forcing, a quantum circuit mapping algorithm that shows a compilation time average speedup of 3.7× compared to the state-of-the-art scalable techniques, reducing the depth of the mapped circuit by 4.7× at the expense of adding 1.3× more SWAP gates. Moreover, the proposed mapping algorithm is adapted and tuned for what is expected to be the next generation of quantum computers, in which different processors are interconnected to increase the total number of qubits, allowing for more complex computations., Authors gratefully acknowledge funding from the European Commission through HORIZON-EIC-2022- PATHFINDEROPEN-01-101099697 (QUADRATURE) and grant HORIZON-ERC-2021-101042080 (WINC). This work has been partially supported by the Spanish Ministry of Science, Innovation and Universities through the Beatriz Galindo program 2020 (BG20-00023), by the European ERDF under grant PID2021-123627OB-C51, and by QCOMMCAT-Planes Complementarios: Comunicacion Cuantica ¿ supported by MICIN with funding from the European Union, NextGenerationEU (PRTR-C17.I1) and Generalitat de Catalunya. The first author gratefully acknowledges the Universitat Politecnica de Catalunya and Banco Santander for ` the financial support of his predoctoral FPI-UPC grant.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

From designing quantum processors to large-scale quantum computing systems

  • GARCIA ALMUDEVER, CARMEN|||0000-0002-3800-2357
  • Willw, Robert
  • Sebastiano, Fabio
  • Haider, Nadia
  • Alarcon, Eduard
[EN] Design, simulation, analysis and verification methodologies are crucial for developing electronic circuits and systems at large. Whereas long-standing EDA software is used in the semiconductor technology, there is no counterpart for quantum computing systems yet. Although the quantum computing community started utilizing and adapting some of the already existing EDA tools, for instance, to design quantum processors and control electronics for driving the qubits, or even to solve some quantum computing design tasks, they do not fully use the expertise gained over the last decades in the field of design automation. Current intermediate-scale quantum computers have been designed in an `adhoc¿ manner with heterogeneous methods and tools. As we are entering the large-scale era, it is timely and key to further adopt EDA methodologies and software for quantum computing. In this paper, we provide an overview on how full-stack quantum computing systems are being implemented nowadays and discuss which the main challenges are for transitioning from this current scenario to a comprehensive framework encompassing full automated system-wide architecting, design, simulation, verification, and test., N.H. expresses gratitude for the support received from Leo Di Carlo and his group, as well as Intel Corporation and IARPA (U.S. Army Research Office Grant No. W911NF16-1-0071) for their contributions to the research funds. C.G.A., F.S. and E.A. acknowledge support from EU, grant HORIZON-EIC-2022-PATHFINDEROPEN-01-101099697. E.A acknowledges support from the QCOMM-CAT-Planes Complementarios: Comunicacion Cuántica and Generalitat de Catalunya. C.G.A also acknowledges support from MICIIN and European ERDF under grant PID2021-123627OB-C51. R.W. s work received funding from the European Research Council (ERC) under the European Union s Horizon 2020 research and innovation program (grant agreement No. 101001318).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Power-aware scheduling strategies for improving performance and energy consumption in heterogeneous multicore platforms

  • Mazón, Ricardo
  • Hassan Mohamed, Houcine|||0000-0002-2038-0587
[EN] Energy consumption is an increasingly critical concern in multicore embedded real-time systems, especially when it is referred to systems that operate under battery constraints as mobile devices. Moreover, these systems must enhance performance such as the execution time while meeting temporal constraints to reduce deadline miss rates. Although Heterogeneous Multi-Processing (HMP) architectures offer opportunities for power optimization using dynamic voltage and frequency scaling (DVFS), standard operating system (OS) governors do not consider benchmark characteristics, memory or CPU usage, and optimal co-runner configurations. This limitation leads to scheduling decisions that trigger early battery depletion or extreme deadline violations under high system utilization. To solve these challenges, this paper proposes a two-level scheduling and DVFS framework. The proposed solution performs profiling benchmark-pair combinations to construct a static co-runner compatibility matrix. Online, a supervisor level (Partitioner) dynamically evaluates global system constraints, tracking remaining battery status and runtime deadline miss rates to adaptively select an heuristic strategy (A_max, B_max, B_eff). In the second level, a Task Scheduler executes tasks based on Earliest Deadline First (EDF) policy, integrated with a power estimation model. This structure translates a nonlinear, NP-hard multi-objective optimization problem into a deterministic, constant-time O(1) runtime overhead per task. Experimental evaluations are conducted on an ODROID XU4 platform containing a Samsung Exynos 5422 heterogeneous multicore processor, based on an ARM big.LITTLE architecture, using a set of embedded real-time benchmarks. The results show that: (i) B_max strategy delivers a suitable trade-off by simultaneously reducing mean energy consumption by about 7% and execution time by around 8%, compared to the OS., at the same time reducing drastically deadline failures by 72.61% under a 50% workload; (ii) the aggressive B_eff configuration, decreases power consumption by 50% in situations with ample slack time; and (iii) the A_max recovery strategy achieves about 29% acceleration, hence preventing system degradation under extreme workload saturation. Finally, integrating these context-aware mechanisms prevents the hardware from performance collapse, maximizing battery longevity while ensuring suitable real-time predictability., This work has been supported by the Spanish Ministerio de Ciencia e Innovación and European ERDF under grants PID2021-123627OB-C51 and PID2024-158682OB-C31, and Generalitat Valenciana under grant CIPROM/2024/007.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Reducing the number of qubits in quantum simulations of one dimensional many-body Hamiltonians

  • Mehrabankar, Somayeh
  • Pérez, Armando
  • Garcia March, Miguel Angel|||0000-0001-7092-838X
  • GARCIA ALMUDEVER, CARMEN|||0000-0002-3800-2357
[EN] We investigate the Ising and Heisenberg models using the block renormalization group method (BRGM), focusing on its behavior across different system sizes. The BRGM reduces the number of spins by a factor of 1/2 (1/3) for the Ising (Heisenberg) model, effectively preserving essential physical features of the model while using only a fraction of the spins. Through a comparative analysis, we demonstrate that as the system size increases, there is an exponential convergence between results obtained from the original and renormalized Ising Hamiltonians, provided the coupling constants are redefined accordingly. Remarkably, for a spin chain with 24 spins, all physical features, including magnetization, correlation function, and entanglement entropy, exhibit an exact correspondence with the results from the original Hamiltonian. The study of the Heisenberg model also shows this tendency, although complete convergence may appear for a size much larger than 24 spins, and is therefore beyond our computational capabilities. The success of BRGM in accurately characterizing the Ising model, even with a relatively small number of spins, underscores its robustness and utility in studying complex physical systems, and facilitates its simulation on current NISQ computers, where the available number of qubits is largely constrained., The authors would like to acknowledge the enlightening discussions with German Sierra and express their gratitude to Ali Najafi (www.linkedin.com/in/ali-najafi-84345546) for his collaboration in numerical calculations. These calculations were performed using the Luis Vives machine at the University of Valencia, Spain. The authors gratefully acknowledge the computer resources at Artemisa, funded by the European Union ERDF and Comunitat Valenciana, as well as the technical support provided by the Instituto de Fisica Corpuscular, IFIC (CSIC-UV). This work has been funded by the Spanish MCIN/AEI/10.13039/501100011033 Grant PID2020-113334GB-I00, Generalitat Valenciana Grant CIPROM/2022/66, the Ministry of Economic Affairs and Digital Transformation of the Spanish Government through the QUANTUM ENIA project call-QUANTUM SPAIN project, and by the European Union through the Recovery, Transformation, and Resilience Plan: NextGenerationEU within the framework of the Digital Spain 2026 Agenda, and by the CSIC Interdisciplinary Thematic Platform (PTI+) on Quantum Technologies (PTI-QTEP+). This project has also received funding from the European Union's Horizon 2020 research and innovation program under Grant Agreement 101086123-CaLIGOLA. M A G-M acknowledges funding from the Spanish Ministry of Education and Professional Training (MEFP) through the Beatriz Galindo program 2018 (BEAGAL18/00203), QuantERA II Cofund 2021 PCI2022-133004, Projects of MCIN with funding from European Union Next GenerationEU (PRTR-C17.I1) and by Generalitat Valenciana, with reference 20220883 (PerovsQuTe) and COMCUANTICA/007 (QuanTwin), and Red Tematica RED2022-134391-T. CGA acknowledges support from the Spanish Ministry of Science, Innovation and Universities through the Beatriz Galindo program 2020 (BG20-00023) and the European ERDF under Grant PID2021-123627OB-C51 and from the QuantERA Grant EQUIP with the Grant Numbers PCI2022-133004, funded by Agencia Estatal de Investigacion, Ministerio de Ciencia e Innovacion, Gobierno de Espana, MCIN/AEI/10.13039/501100011033, and by the European Union 'NextGenerationEU/PRTR'
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Implementing Quantum Polar Codes in a Superconducting Processor: From State Preparation to Decoding [Focus: Quantum Software and Its Engineering]

  • Kurniawan, Handy
  • Savin, Valentin
  • Garcia Herrero, Francisco
  • GARCIA ALMUDEVER, CARMEN|||0000-0002-3800-2357
[EN] A full-stack software framework for implementing quantum polar codes is presented, integrating a noise-aware compilation approach that optimizes resource usage by combining quantum and classical software. Experimental results demonstrate significant improvements in logical state preparation rates, resulting in reduced comsumption of both quantum and classical resources., This work was supported by the QuantERA project EQUIP (Grants PCI2022-133004 and PCI2022-132922, funded by the AEI, Ministerio de Ciencia e Innovacion, Gobierno de Espana, MCIN/AEI/10. 13039/501100011033, and ANR-22-QUA2-0005-01, funded by the Agence Nationale de la Recherche, France), and by the European Union "Next Generation EU/PRTR". This research is part of the project PID2023-147059OB-I00 funded by MCIU/AEI/10 .13 039/50110 0011033/FEDER, UE. Handy Kurniawan acknowledges support from the Comunidad de Madrid under Grant PIPF-2023/COM-30051. Carmen G. Almudever acknowledges support from the Spanish Ministry of Science, Innovation, and Universities through the Beatriz Galindo program 2020 (BG20-00023) and the European ERDF PID2021-123627OB-C51. We acknowledge the use of IBM Quantum services for this work. The views expressed are those of the authors and do not reflect the official policy or position of IBM or the IBM Quantum team. This article has supplementary downloadable material available at https://doi.org/10.1109/MS.2025.3569447, provided by the authors.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Análisis de Interferencias en la Red de Interconexión entre Aplicaciones Exascale

  • Baviera-Guaita, Alejandro
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • Kaidanov, Max
  • Duro, José
[ES] En los supercomputadores exascala, la red de interconexion se convierte en uno de los componentes mas criticos. Las topologias tradicionales no satisfacen las necesidades de prestaciones y consumo de estos sistemas y por ello se utilizan topolog¿as mas novedosas con conmutadores de alto grado como la Dragonfly. Sin embargo, esta topologia tambien presenta sus problemas de aislamiento de los trabajos que comparten el sistema. Los trabajos independientes que se ejecutan concurrentemente se ven obligados a compartir recursos generando interferencias entre ellos en las comunicaciones a traves de la red. Esta interferencia se magnifica aun mas por el hecho de que las redes Dragonfly funcionan mejor bajo trafico aleatorio, lo que pone de manifiesto la importancia del analisis junto a la interferencia de las politicas de mapeo de aplicaciones a nodos de procesamiento en el sistema. En este articulo analizamos en la topolog¿a Dragonfly y en el contexto de la arquitectura de red BXI la interferencia sufrida por aplicaciones reales del dominio HPC considerando varias estrategias de mapeo de estas aplicaciones en los nodos de procesamiento. Para ello analizamos dos escenarios, un escenario en el que una aplicacion de las seleccionadas en el proyecto europeo RED-SEA se ejecuta junto a trafico sintetico de distinta intensidad y un segundo escenario donde dos aplicaciones reales de las de referencia del proyecto se ejecutan simultaneamente en la maquina. Los resultados muestran que la politica de mapeo tiene mucha influencia en el tiempo de ejecucion de la aplicacion y en la interferencia ejercida entre aplicaciones., El presente trabajo ha sido financiado por la Comision Europea (programa Horizon 2020, call H2020- JTI-EuroHPC-2019-1 - grant agreement ID: 955776) y Ministerio de Ciencia e Innovacion de España y Fondos Europeos ERDF (proyectos PCI2021-121934, PCI2021-121976, TED2021-130233B-C32, PID2021- 123627OB-C51.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Cloud White: Detecting and Estimating QoS Degradation of Latency-Critical Workloads in the Public Cloud

  • Pons-Escat, Lucía|||0000-0002-4582-7744
  • Feliu-Pérez, Josué|||0000-0003-3017-4266
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Pons Terol, Julio|||0000-0002-5654-6753
  • Huang, Chaoyi
[EN] The increasing popularity of cloud computing has forced cloud providers to build economies of scale to meet the growing demand. Nowadays, data-centers include thousands of physical machines, each hosting many virtual machines (VMs), which share the main system resources, causing interference that can significantly impact on performance. Frequently, these data-centers run latency-critical workloads, whose performance is determined by tail latency, which is very sensitive to the interference of co-running workloads. To prevent QoS violations, cloud providers adopt overprovisioning strategies but they reduce the server utilization and increase the costs. A mechanism that accurately estimates performance degradation dynamically in a production system would allow cloud providers to improve the servers' utilization. In this work we propose Cloud White, an approach that is able to detect the inter-VM interference in scenarios with multiple co-located latency-critical VMs and estimate the performance degradation using multi-variable regression models. Unlike previous proposals, Cloud White is built taking into account the limitations of a public cloud production system. Experimental results show that Cloud White is able to estimate performance degradation with a small overall prediction error of 5%., This work has been supported by Huawei Cloud, and in part by Spanish Ministerio de Universidades under grant FPU18/01948, and by Spanish Ministerio de Universidades and European ERDF under grants RTI2018-098156-B-C51 and PID2021-123627OB-C51. Funding for open access charge: CRUE-Universitat Politec-nica de Valencia.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación 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
Publicación 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
Publicación Contribución de congreso (ConferenceOutput)

Mapping quantum algorithms to multi-core quantum computing architectures

  • Ovide-González, Anabel
  • GARCIA ALMUDEVER, CARMEN|||0000-0002-3800-2357
  • Rodrigo, Santiago
  • Bandic, Medina
  • Van Someren, Hans
  • Feld, Sebastian
  • Abadal, Sergi
  • Alarcón, Eduard
[EN] Current monolithic quantum computer architectures have limited scalability. One promising approach for scaling them up is to use a modular or multi-core architecture, in which different quantum processors (cores) are connected via quantum and classical links. This new architectural design poses new challenges such as the expensive inter-core communication. To reduce these movements when executing a quantum algorithm, an efficient mapping technique is required. In this paper, a detailed critical discussion of the quantum circuit mapping problem for multi-core quantum computing architectures is provided. In addition, we further explore the performance of a mapping method, which is formulated as a partitioning over time graph problem, by performing an architectural scalability analysis., We acknowledge support from EU, grant HORIZONERC-101042080 (S.A.) and grant HORIZON-EIC-2022- PATHFINDEROPEN-01-101099697 (S.A., E.A and C.G.A.), from project PRG 946 funded by the Estonian Research Council (A.O.), from MICIIN and European ERDF under grant PID2021-123627OB-C51 (C.G.A.) and from MICIIN with funding from European Union NextGenerationEU(PRTRC17.I1) and by Generalitat de Catalunya (E.A.)
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

On the use of calibration data in error-aware compilation techniques for NISQ devices

  • Kurniawan, Handy
  • Cuomo, Daniele
  • Garcia Herrero, Francisco
  • Rodríguez-Soriano, Laura|||0000-0002-2074-0466
  • GARCIA ALMUDEVER, CARMEN|||0000-0002-3800-2357
[EN] Reliably executing quantum algorithms on noisy intermediate-scale quantum (NISQ) devices is challenging, as they are severely constrained and prone to errors. Efficient quantum circuit compilation techniques are therefore crucial for overcoming their limitations and dealing with their high error rates. These techniques consider the quantum hardware restrictions, such as the limited qubit connectivity, and perform some transformations to the original circuit that can be executed on a given quantum processor. Certain compilation methods use error information based on calibration data to further improve the success probability or the fidelity of the circuit to be run. However, it is uncertain to what extent incorporating calibration information in the compilation process can enhance the circuit performance. For instance, considering the most recent error data provided by vendors after calibrating the processor might not be functional enough as quantum systems are subject to drift, making the latest calibration data obsolete within minutes. In this paper, we explore how different usage of calibration data impacts the circuit fidelity, by using several compilation techniques and quantum processors (IBM Perth and Brisbane). To this aim, we implemented a framework that incorporates some of the state-of-the-art noise-aware and non-noise-aware compilation techniques and allows the user to perform fair comparisons under similar processor conditions. Our experiments yield valuable insights into the effects of noise-aware methodologies and the employment of calibration data. The main finding is that pre-processing historical calibration data can improve fidelity when real-time calibration data is not available due to factors such as cloud service latency and waiting queues between compilation and execution on the quantum backend., This work is supported by the QuantERA grant EQUIP with the grant numbers PCI2022-133004 and PCI2022- 132922, funded by Agencia Estatal de Investigacion, Ministerio de Ciencia e Innovacion, Gobierno de España, MCIN/AEI/10.13039/501100011033, and by the European Union NextGenerationEU/PRTR . CGA acknowledges funding from the Spanish Ministry of Science, Innovation and Universities through the Beatriz Galindo program 2020 (BG20- 00023) and from the European ERDF under grant PID2021- 123627OB-C51. We acknowledge the use of IBM Quantum services for this work. The views expressed are those of the authors and do not reflect the official policy or position of IBM or the IBM Quantum team.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación 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.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

WAPA: A Workload-Agnostic CPI-Based Thread-to-Core Allocation Policy

  • Navarro-Edo, Marta|||0000-0002-4334-460X
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • Pallardó-Julià, Vicent
[EN] Simultaneous multithreading (SMT) processors improve system throughput by sharing core resources among the threads running on the same core. However, intra-core interference can cause co-running applications to degrade each other¿s performance significantly. To address this issue, some approaches have focused on balancing contention at the core shared resources (e.g. the shared L1 data cache). A key advantage of these approaches is that they are workload-agnostic. Other approaches improve the previous ones by modeling the inter-application interference across the intra-core shared resources. Unfortunately, these approaches require off-line model training for specific workloads. This paper presents WAPA, a CPI-based thread-to-core allocation approach that incorporates the best of both worlds. WAPA is a workloadagnostic policy that implicitly accounts for inter-thread interference across all the shared resources by leveraging the CPI. The proposed approach relies on the optimal transport (OT) theory, a mathematical theory to dynamically select symbiotic pairs of applications. Experimental results in an Intel Xeon show that WAPA outperforms the default Linux scheduler on average by 8.4% in IPC in workloads dominated by cache and main memory latencies, while performance gains of existing approaches are below 3.2%., This work has been partially 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
Publicación Contribución de congreso (ConferenceOutput)

Diseño de un controlador de memoria principal híbrido para aplicaciones de Big Data

  • Lurbe, Manel
  • Avargues, Miguel A.
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
[ES] La tecnologia de memoria DRAM ha sido la dominante en las implementaciones del subsistema de memoria. En los ultimos años, e impulsada principalmente por las enormes demandas de memoria de las aplicaciones de big data, la tecnologia NVRAM ha emergido como una tecnologia de memoria m¿as densa, permitiendo el diseño de nuevas jerarquias de memoria hibridas DRAM/NVRAM que combinan multiples tecnologias de memoria para equilibrar capacidad, latencia, coste y persistencia. En el diseño de jerarquias de memoria hibridas se est¿an aplicando dos enfoques principales: el enfoque de espacio de direcciones hibrido DRAM/NVRAM, que confia en el programador o el sistema operativo para elegir la tecnologia de memoria en la que debe almacenarse cada pagina de memoria; y el enfoque de espacio de direcciones (unico) NVRAM, en el que se necesita una tecnologia mas rapida (por ej., DRAM convencional) que actue como cache de la NVRAM para aumentar las prestaciones. Este enfoque presenta retos arquitectonicos como la organizacion de metadatos. En contraste con los enfoques existentes, este trabajo propone un controlador de memoria que aprovecha tecnologias de memoria recientes como eDRAM y MRAM para mejorar las prestaciones de la lenta memoria NVRAM cuando actua como memoria principal. La propuesta consiste en una jerarquia de cache de dos niveles en el controlador de memoria: una cache de sectores SRAM y una cache (x)RAM. La cache (x)RAM, mucho mas densa, ayuda a reducir significativamente el numero de accesos a la NVRAM. Los resultados experimentales demuestran que la implementaci¿on de la cache (x)RAM con eDRAM o MRAM es el metodo mas eficaz. Ademas, la eDRAM es capaz de mejorar la penalizacion en caso de fallo de la cache SRAM hasta en un 50 % y un 80 %, y las prestaciones globales del sistema en un 15 % y un 23 %., Este trabajo ha sido financiado por el Ministerio de Ciencia e Innovacion español y el FEDER europeo mediante las subvenciones PID2021-123627OBC51 y TED2021-130233B-C32. Manel Lurbe cuenta con el apoyo de Ayudas para Contratos predoctorales UPV - subprograma 1 (PAID-01-20).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Lightcone Bounds for Quantum Circuit Mapping via Uncomplexity

  • Steinberg, Matthew
  • Bandic, Medina
  • Szkudlarek, Sacha
  • Sarkar, Aritra
  • Feld, Sebastian
  • GARCIA ALMUDEVER, CARMEN|||0000-0002-3800-2357
[EN] Efficiently mapping quantum circuits onto hardware is integral for the quantum compilation process, wherein a circuit is modified in accordance with a quantum processor¿s connectivity. Many techniques currently exist for solving this problem, wherein SWAP-gate overhead is usually prioritized as a cost metric. We reconstitute quantum circuit mapping using tools from quantum information theory, showing that a lower bound, which we dub the lightcone bound, emerges for a circuit executed on hardware. We also develop an initial placement algorithm based on graph similarity search, aiding us in optimally placing circuit qubits onto a device. 600 realistic benchmarks using the IBM Qiskit compiler and a brute-force method are then tested against the lightcone bound, with results unambiguously verifying the veracity of the bound, while permitting trustworthy estimations of minimal overhead in near-term realizations of quantum algorithms. This work constitutes the first use of quantum circuit uncomplexity to practically-relevant quantum computing., Quantum circuit mapping; Quantum compilation; Lightcone bound; Graph similarity; Qubit placement; Quantum information theory
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Interaction graph-based profiling of quantum benchmarks for improving quantum circuit mapping techniques

  • Bandic, Medina
  • Feld, Sebastian
  • GARCIA ALMUDEVER, CARMEN|||0000-0002-3800-2357
[EN] To execute quantum circuits on a quantum processor, they must be modifed to meet the physical constraints of the quantum device. This process, called quantum circuit mapping, results in a gate/circuit depth overhead that depends on both the circuit properties and the hardware constraints, being the limited qubit connectivity a crucial restriction. In this paper, we propose to extend the characterization of quantum circuits by including qubit interaction graph properties using graph theorybased metrics in addition to previously used circuit-describing parameters. This approach allows for an in-depth analysis and clustering of quantum circuits and a comparison of performance when run on diferent quantum processors, aiding in developing better mapping techniques. Our study reveals a correlation between interaction graph-based parameters and mapping performance metrics for various existing confgurations of quantum devices. We also provide a comprehensive collection of quantum circuits and algorithms for benchmarking future compilation techniques and quantum devices., This work has been partially supported by the Spanish Ministerio de Ciencia e Innovación and European ERDF under grant PID2021-123627OB-C51 and by the QuantERA grant EQUIP, by the Ministerio de Ciencia e Innovación and Agencia Estatal de Investigación, MCIN/AEI/10.13039/501100011033, and by the European Union NextGenerationEU/PRTR (CGA).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Rebasing Microarchitectural Research with Industry Traces

  • Feliu-Pérez, Josué|||0000-0003-3017-4266
  • Perais, Arthur
  • Jiménez, Daniel A.
  • Ros, Alberto
[EN] Microarchitecture research relies on performance models with various degrees of accuracy and speed. In the past few years, one such model, ChampSim, has started to gain significant traction by coupling ease of use with a reasonable level of detail and simulation speed. At the same time, datacenter class workloads, which are not trivial to set up and benchmark, have become easier to study via the release of hundreds of industry traces following the first Championship Value Prediction (CVP-1) in 2018. A tool was quickly created to port the CVP-1 traces to the ChampSim format, which, as a result, have been used in many recent works. In this paper, we revisit this conversion tool and find that several key aspects of the CVP-1 traces are not preserved by the conversion. We therefore propose an improved converter that addresses most conversion issues as well as patches known limitations of the CVP-1 traces themselves. We evaluate the impact of our changes on two commits of ChampSim, with one used for the first Instruction Championship Prefetching (IPC-1) in 2020. We find that the performance variation stemming from higher accuracy conversion is significant., This project has received funding from the European Research Council (ERC) under the European Union s Horizon 2020 research and innovation programme (grant agreement No 819134), from the MCIN/AEI/10.13039/501100011033/ and the ERDF A way of making Europe , EU (grants PID2021-123627OB-C51 and PID2022-136315OB-I00) and the European Union NextGenerationEU/PRTR (grants RYC2021- 030862-I and TED2021-130233B-C33/C32), from the National Science Foundation (grants CNS-1938064 and CCF-1912617), as well as generous gifts from Intel. Portions of this research were conducted with the advanced computing resources provided by Texas A&M High Performance Research Computing.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Tesis doctoral (DoctoralThesis)

Optimizing Prefetch with Machine Learning and Analyzing its Impact on the Performance of Deep Memories Hierarchies

  • Lurbe Sempere, Manel
[ES] La creciente presión sobre los recursos compartidos de la jerarquía de memoria en los sistemas de cómputo modernos se ha convertido en un factor crítico que limita el rendimiento. La proliferación de aplicaciones concurrentes, impulsada por cargas de trabajo de aprendizaje automático y entornos virtualizados en la nube, ha transformado la concepción tradicional de las jerarquías de memoria y acentuado la necesidad de ampliar la capacidad de la memoria principal. Sin embargo, la memoria DRAM convencional enfrenta restricciones físicas y tecnológicas que comprometen su escalabilidad, lo que ha impulsado la incorporación de memorias no volátiles (NVRAM) como alternativa emergente. A pesar de sus ventajas en densidad, la mayor latencia de acceso de la NVRAM plantea nuevos desafíos que requieren rediseñar y optimizar los mecanismos de gestión de memoria. En este contexto, los mecanismos de prefetching resultan fundamentales para mitigar las penalizaciones de latencia y aprovechar plenamente las nuevas tecnologías de memoria. No obstante, la configuración de prefetchers en arquitecturas multinivel plantea tensiones significativas, dado que su carácter especulativo incrementa el consumo de ancho de banda y puede deteriorar el rendimiento en sistemas con alta concurrencia. Así, la selección adecuada de configuraciones de prefetching requiere un equilibrio entre las características de acceso de las aplicaciones y la eficiencia en el uso del ancho de banda compartido. De manera complementaria, la integración de memorias NVRAM en la jerarquía introduce profundos cambios estructurales. La menor velocidad de estas memorias en comparación con la DRAM exige el diseño de nuevas organizaciones de caché y controladores de memoria capaces de mitigar sus limitaciones de latencia y preservar el rendimiento de los procesadores de altas prestaciones. Esta tesis aborda de forma conjunta dos problemáticas clave en la evolución de las jerarquías de memoria: (i) la selección óptima de configuraciones de prefetching en cargas de trabajo multiprogramadas mediante técnicas basadas en redes neuronales, y (ii) el diseño de jerarquías profundas con NVRAM orientadas a aumentar la capacidad de memoria principal sin comprometer el rendimiento., [EN] The increasing pressure on shared resources within the memory hierarchy of modern computing systems has become a critical performance-limiting factor. The proliferation of concurrent applications, driven by machine learning workloads and cloud-based virtualized environments, has reshaped the traditional conception of memory hierarchies and highlighted the need to expand main memory capacity. However, conventional DRAM faces physical and technological constraints that hinder its scalability, fostering the integration of non-volatile memories (NVRAM) as an emerging alternative. Despite their advantages in density, NVRAM technologies introduce higher access latency, which demands new strategies for memory management and optimization. In this context, prefetching mechanisms play a fundamental role in mitigating latency penalties and fully exploiting the potential of emerging memory technologies. Nevertheless, the configuration of prefetchers in multilevel architectures poses significant challenges, as their speculative nature increases bandwidth consumption and can degrade performance in highly concurrent systems. Therefore, the optimal selection of prefetching configurations requires balancing application-specific access patterns with efficient shared bandwidth utilization. Complementarily, the integration of NVRAM into the hierarchy introduces profound structural changes. The slower speed of these memories compared to DRAM necessitates the design of new cache organizations and memory controllers to mitigate latency limitations and sustain high-performance processor efficiency. This dissertation %jointly addresses two key challenges in the evolution of memory hierarchies: (i) the optimal selection of prefetching configurations in multiprogrammed workloads through neural network-based techniques, and (ii) the design of deep NVRAM hierarchies aimed at increasing main memory capacity without compromising system performance., [CA] La creixent pressió sobre els recursos compartits dins de la jerarquia de memòria dels sistemes de computació moderns s'ha convertit en un factor crític que limita el rendiment. La proliferació d'aplicacions concurrents, impulsada per les càrregues de treball d'aprenentatge automàtic i els entorns virtualitzats al núvol, ha transformat la concepció tradicional de les jerarquies de memòria i ha posat de manifest la necessitat d'ampliar la capacitat de la memòria principal. No obstant això, la DRAM convencional afronta restriccions físiques i tecnològiques que en dificulten l'escalabilitat, fet que ha propiciat la incorporació de memòries no volàtils (NVRAM) com a alternativa emergent. Tot i els seus avantatges en densitat, les NVRAM introdueixen una latència d'accés superior, la qual cosa exigeix noves estratègies de gestió i optimització de la memòria. En aquest context, els mecanismes de prefetching són fonamentals per a mitigar les penalitzacions de latència i aprofitar plenament el potencial de les noves tecnologies de memòria. Tanmateix, la configuració dels prefetchers en arquitectures multinivell planteja reptes importants, atés que el seu caràcter especulatiu incrementa el consum d'ample de banda i pot degradar el rendiment en sistemes amb alta concurrència. Per tant, la selecció òptima de configuracions de prefetching requereix equilibrar els patrons d'accés específics de les aplicacions amb una utilització eficient dels recursos compartits. De manera complementària, la integració de NVRAM en la jerarquia introdueix canvis estructurals profunds. La menor velocitat d'aquestes memòries en comparació amb la DRAM obliga a dissenyar noves organitzacions de memòria cau i nous controladors de memòria capaços de mitigar les seues limitacions de latència i mantindre l'eficiència dels processadors d'altes prestacions. Aquesta tesi aborda conjuntament dos reptes clau en l'evolució de les jerarquies de memòria: (i) la selecció òptima de configuracions de prefetching en càrregues de treball multiprogramades mitjançant tècniques basades en xarxes neuronals, i (ii) el disseny de jerarquies profundes amb NVRAM orientades a augmentar la capacitat de la memòria principal sense comprometre el rendiment del sistema., This work has been supported by the Spanish Ministerio de Ciencia e Innovación and European ERDF under grants PID2021-123627OB-C51, PID2024-158682OB-C31, by MCIN/AEI/10.13039/501100011033 and European Union NextGenerationEU/PRTR under grant TED2021-130233B-C32, and Generalitat Valenciana under grant CIPROM/2024/007. Manel Lurbe is supported by Ayudas Contratos predoctorales UPV-subprograma 1 (PAID-01-20).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

WAPA: A Microarchitecture- and Workload-Agnostic Universal SMT Scheduler

  • Navarro-Edo, Marta|||0000-0002-4334-460X
  • Pons-Escat, Lucía|||0000-0002-4582-7744
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • Pallardó-Julià, Vicent
[EN] This paper presents WAPA, a thread-to-core allocation policy for SMT processors grounded in OT theory, and demonstrates its effectiveness in achieving two key design goals: microarchitecture agnosticism and workload agnosticism. WAPA is implemented on both Intel Xeon and ARM ThunderX2 platforms without requiring modifications, leveraging universally available performance counters. Experimental results show that WAPA consistently outperforms both the default Linux scheduler and state-of-the-art model-based approaches, achieving up to 110% improvement in turnaround time on Intel and over 30% average gains on ARM. Furthermore, we show WAPA is extensible to support alternative cost metrics beyond CPI, such as energy consumption or fairness. These findings position WAPA as a practical and portable solution for integration into modern operating system schedulers., The work of Marta Navarro was supported by Generalitat Valencianathrough "Subvenciones para la contratacion de personal investigador pre-doctoral" under Grant CIACIF/2021/413. This work was supported in partby MICIU/AEI/10.13039/501100011033/ FEDER, UE under Grant PID2024-158682OB-C31 and Grant PID2021-123627OB-C51 and in part by GeneralitatValenciana under Grant CIPROM/2024/7.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Optimization of One-to-Many Communication Primitives for Dragonfly Topologies

  • Duro, José
  • Castelló, Adrián|||0000-0002-8576-8451
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • Quintana-Ortí, Enrique S.|||0000-0002-5454-165X
[EN] Collective communication primitives (CCPs), such as multicast and broadcast, are essential for many parallel and distributed applications. In response, this study compares a topology-oblivious algorithm underlying the implementation of CCPs in standard instances of MPI with two topology-aware implementations, based on the LLF and GLF algorithms, and an ideal hardware-assisted approach. By using real scientific applications, instead of synthetic traffic, our study reveals workloaddependent performance variations among CCP implementations; and highlights the importance of CCP algorithm selection in optimizing application performance in supercomputing environments., This work is supported by RED-SEA, a project that receives funding from the European HPC JU under grant agreement No 955776 and the research and innovation programme of Spain (PCI2021-121934); and Ministerio de Ciencia e Innovacion y Fondos ERDF PID2021-123627OB-C51.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

PANES(Parejas afiNES): Estrategia de mejora de prestaciones en procesadores SMT de ARM

  • Navarro-Edo, Marta|||0000-0002-4334-460X
  • Feliu-Pérez, Josué|||0000-0003-3017-4266
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
[ES] Este artículo presenta PANES, una estrategia sencilla que asigna dinámicamente hilos a núcleos de un procesador SMT en función de su comportamiento dinámico en tiempo de ejecución., Este trabajo ha sido financiado por el Ministerio de Ciencia e Innovaci¿on junto al FEDER europeo a trav¿es de los proyectos PID2021-123627OBC51, TED2021-130233B-C3 y AICO/2021/266. Marta Navarro tiene la beca Subvenciones para la contrataci¿on de personal investigador predoctoral de la Generalitat Valenciana (ACIF2022). Josu¿e Feliu cuenta con el apoyo del contrato RYC2021-030862-I financiado por MCIN/AEI /10.13039/501100011033 y por la Uni¿on Europea NextGenerationEU/PRTR.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Modelado de una Arquitectura de Monitorización de Prestaciones con el Simulador Gem5

  • Pecino-Fernández, Manuel Antonio
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
[ES] Los simuladores son una herramienta esencial en la investigación de la Arquitectura de Computadores, ya que proporcionan una solución ágil y económica para el análisis preciso y eficiente de propuestas estructurales y microarquitectónicas para el diseño de CPU. Una de sus principales ventajas de estas herramientas es la capacidad de análisis de las prestaciones del procesador, permitiendo obtener un conocimiento profundo sobre el rendimiento de las etapas del pipeline y el estado de la máquina. Sin embargo, los simuladores actuales carecen de un modelado preciso de las Performance Monitoring Units (PMU) presentes en procesadores reales, lo que limita su capacidad para realizar ciertos análisis avanzados de rendimiento como el análisis Top-Down propuesto por Intel. En el presente trabajo se realiza la implementación de una PMU para el simulador gem5, incluyendo la capacidad de monitorizar aplicaciones ejecutándose en procesadores con soporte a multihilo SMT (Simultaneous Multi-Threading). Los resultados muestran la viabilidad y utilidad de esta nueva funcionalidad, que permite realizar análisis Top-Down de cargas de trabajo SPEC CPU tanto monohilo como multihilo., Este trabajo ha sido financiado por el Ministerio de Ciencia e Innovacion junto al FEDER europeo a traves de los proyectos PID2021-123627OB-C51 y TED2021-130233B-C32.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Stratus: A Hardware/Software Infrastructure for Controlled Cloud 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
  • Huang, Chaoyi
[EN] Cloud systems deploy a wide variety of shared resources and host a large number of tenant applications. To perform cloud research, a small experimental platform is commonly used, which hides the huge system complexity and provides flexibility. Despite being simpler, this platform should include the main cloud system components (hardware and software) to provide representative results. A wide set of platforms have spread in recent years; however, most of them only include a major cloud component or lack the deployment of virtual machines (VMs) to provide isolation. This paper presents Stratus, an experimental platform that is currently being used to carry out cloud research. To the best of our knowledge, Stratus is the only platform that jointly provides three main features: uses VMs to isolate tenant applications, deploys the three types of cloud nodes (server, client, and storage), and manages all main shared system resources (CPUs, LLC space, memory, network, and disk bandwidth). Moreover, Stratus implements a software manager to ease the research and aid the design of QoS-aware policies. The manager integrates three main functionalities: management and control of the execution of VMs and running applications, monitoring of hardware performance counters and system resource utilization, and partitioning of the main shared system resources by using technologies available in commercial processors., This work has been partially supported by Huawei Cloud, by the Spanish Ministerio de Universidades under the grant FPU18/01948, and by the Spanish Ministerio de Ciencia e Innovacion and European ERDF under grants PID2021- 123627OB-C51 and TED2021-130233B-C32, and by Generalitat Valenciana under Grant AICO/2021/266.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Desarrollo de una infrastructura para la investigación en cloud computing

  • 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
[ES] Los sistemas en la nube despliegan una amplia variedad de recursos compartidos y albergan un gran número de aplicaciones. Para realizar investigación dentro de este ámbito, se suele emplear un sistema reducido que oculta la enorme complejidad de los sistemas reales y proporciona flexibilidad para gestionar los recursos. A pesar de ser más sencillos y a menor escala, estos sistemas experimentales deben incluir los principales componentes (hardware y software) de los sistemas de computación en la nube o cloud computing para proporcionar resultados representativos. Este trabajo presenta Stratus, una plataforma experimental que se está utilizando actualmente para llevar a cabo investigación en cloud computing. A diferencia de otras plataformas existentes, Stratus es la única plataforma que ofrece conjuntamente todas las características principales de estos sistemas: uso de máquinas virtuales (MV) para ejecutar las aplicaciones de los usuarios, despliegue de los tres tipos de nodos de la nube (servidor, cliente y almacenamiento) y gestión de los principales recursos compartidos del sistema (CPU, espacio de caché, ancho de banda de memoria, red y disco). Además, Stratus implementa un gestor de software para facilitar la investigación y ayudar al diseño de políticas orientadas a mejorar la calidad de servicio., Este trabajo ha sido financiado por Huawei Cloud, por el Ministerio de Universidades mediante la beca FPU18/01948, por el Ministerio de Ciencia e Innovacion junto al FEDER europeo a traves de los proyectos PID2021-123627OB-C51 y TED2021-130233BC32.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Contribución de congreso (ConferenceOutput)

Diseño de Políticas de Asignación de Hilos en Procesadores ARM SMT Basadas en Categorías

  • Navarro-Edo, Marta|||0000-0002-4334-460X
  • Feliu-Pérez, Josué|||0000-0003-3017-4266
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • Gómez Requena, María Engracia|||0000-0003-1466-4118
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
[ES] Los servidores HPC modernos integran procesadores simultaneous multithreading (SMT). Estos procesadores comparten los recursos internos del nucleo entre diversos hilos de ejecucion, lo cual les permite alcanzar mayores prestaciones. Sin embargo, esto introduce interferencias entre los hilos que se ejecutan en el mismo nucleo. Esta interferencia intranucleo pone en peligro la mejora de prestaciones y puede aumentar el tiempo de ejecucion. Las pol¿ticas de asignacion de hilos a nucleos (Thread-to-Core Allocation, T2C ) abordan este problema en dos pasos: (i) construir una performance stack por aplicacion usando contadores hardware y (ii) predecir las parejas afines de aplicaciones y asignarlas al mismo nucleo. Este art¿culo presenta la Instructions and Stall Cycles (ISC) stack, un nuevo enfoque diseñado para procesadores ARM resolviendo las limitaciones de la PMU de estos procesadores. Estas stacks de ISC se utilizan en un modelo de prediccion para determinar las parejas afines que ayuden a mejorar las prestaciones. Este trabajo presenta el diseño y analisis de SYNPA, una familia de pol¿ticas T2C guiada por modelos de prediccion de prestaciones basados en stacks de ICS. Los experimentos demuestran que SYNPA4, la mejor variante, reduce el tiempo de ejecucion en un 38 % frente a Linux, logrando 3× las mejoras de pol¿ticas del estado del arte. SYNPA4 se puede aplicar a otros procesadores SMT siguiendo la metodolog¿a presentada en este trabajo., Este trabajo ha sido parcialmente financiado por por el Ministerio de Ciencia e Innovacion junto al FEDER europeo a traves de los proyectos PID2021-123627OB-C51 y TED2021-130233B-C32. Marta Navarro tiene la beca de Subvenciones para la contratacion de personal investigador predoctoral by CIACIF/2021/413. Josue Feliu cuenta con el apoyo del contrato RYC2021-030862-I financiado por MCIN/AEI /10.13039/501100011033 y por la Union Europea NextGenerationEU/PRTR.