Pasar al contenido principal

ARQUITECTURA DE COMPUTADORES DE ALTAS PRESTACIONES

PID2023-146511NB-I00

•
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 2023
Plan Estatal 2021-2023 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario BARCELONA SUPERCOMPUTING CENTER CENTRO NACIONAL DE SUPERCOMPUTACION

Resultados relacionados

Resultados totales (Incluyendo duplicados): 23
Encontrada(s) 1 página(s)

Dipòsit Digital de Documents de la UAB
Publicación Artículo científico (JournalArticle)

BIMSA, accelerating long sequence alignment using processing-in-memory

  • Alonso-Marín, Alejandro
  • Fernandez, Ivan
  • Aguado-Puig, Quim|||0000-0003-4871-3192
  • Gómez-Luna, Juan|||0000-0002-6514-1571
  • Marco-Sola, Santiago|||0000-0001-7951-3914
  • Mutlu, Onur
  • Moreto, Miquel|||0000-0002-9848-8758
Recent advances in sequencing technologies have stressed the critical role of sequence analysis algorithms and tools in genomics and healthcare research. In particular, sequence alignment is a fundamental building block in many sequence analysis pipelines and is frequently a performance bottleneck both in terms of execution time and memory usage. Classical sequence alignment algorithms are based on dynamic programming and often require quadratic time and memory with respect to the sequence length. As a result, classical sequence alignment algorithms fail to scale with increasing sequence lengths and quickly become memory-bound due to data-movement penalties. Processing-In-Memory (PIM) is an emerging architectural paradigm that seeks to accelerate memory-bound algorithms by bringing computation closer to the data to mitigate data-movement penalties. This work presents BIMSA (idirectional n-emory equence lignment), a PIM design and implementation for the state-of-the-art sequence alignment algorithm BiWFA (Bidirectional Wavefront Alignment), incorporating new hardware-aware optimizations for a production-ready PIM architecture (UPMEM). BIMSA supports aligning sequences up to 100K bases, exceeding the limitations of state-of-the-art PIM implementations. First, BIMSA achieves speedups up to 22.24× (11.95× on average) compared to state-of-the-art PIM-enabled implementations of sequence alignment algorithms. Second, achieves speedups up to 5.84× (2.83× on average) compared to the highest-performance multicore CPU implementation of BiWFA. Third, BIMSA exhibits linear scalability with the number of compute units in memory, enabling further performance improvements with upcoming PIM architectures equipped with more compute units and achieving speedups up to 9.56× (4.7× on average).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Leveraging HLS to Design a Versatile &amp, High-Performance Classic McEliece Accelerator

  • Kostalabros, Ioannis-Vatistas
  • Ribes, Jordi
  • Carril, Xavier
  • Farràs, Oriol
  • Moreto, Miquel
  • Hernández Luz, Carles|||0000-0003-1192-6223
[EN] By harnessing fundamental quantum properties, a large-scale quantum computer could undermine currently deployed public-key algorithms. The post-quantum, code-based cryptosystem Classic McEliece (CM)addresses this security concern. However, its large public key size (up to 1.3 MB) poses various hardware implementation challenges. In this article, we focus on the high memory bandwidth requirements of the CM encoding function, in the context of heterogeneous CPU-FPGA devices. More concretely, we target the acceleration of public-key loading and processing from any globally shared or accelerator-private memory system. We present a novel and constant-time accelerator eEnc that exploits the elevated parallelization potential of FPGA devices to yield high-performance results. Our accelerator implements the encoding and the random error vector generation functions, which comprise the main computational load of Encapsulation. Two accelerator design variants are introduced, providing different hardware tradeoffs. Regarding intra-accelerator data communication, and unlike other state-of-the-art (SOTA) works, we combine a streaming protocol with task-level parallelization to remove the need to store the public key in accelerator-private memories. Our proposed design shows new record execution times over its SOTA counterparts, ranging on average from 3.5x up to 7.7x across the five security level parameter sets. Our end-to-end implementation in a Zynq SoC shows an average speedup of 2.2x compared to a 64-bit vectorized CM software-baseline. The elevated logic resource consumption, characteristic of HLS designs, can be readily adjusted with a performance tradeoff., This work has been partially supported by the European HiPEAC Network of Excellence, by the Spanish Ministry of Science and Innovation MCIN/AEI 10.13039/501100011033 (contracts ACITHEC PID2021-124928NB-I00, PID2019-107255GB-C21 and "Ramon y Cajal" fellowship No. RYC2020-030685-I) , by the Generalitat de Catalunya (contract 2021-SGR-00763, 2021-SGR-00115 and 2023-146511NB-I00) , by the Agency for Management of University and Research Grants (AGAUR) of the Government of Catalonia under "Ajuts per a la contractacio de personal investigador novell" (fellowships No. 2019FI B01274 and 2024 FI-1 00520) , by the European Union within the framework of the ERDF of Catalonia 2014-2020 under the DRAC project [001-P-001723] , and by the project HERMES funded by the European NextGenerationEU/PRTR via INCIBE.
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Delegato: Locality-aware atomic memory operations on chiplets

  • Soria Pardos, Víctor
  • Armejach Sanosa, Adrià|||0000-0003-2869-668X
  • Mück, Tiago
  • Suárez Gracía, Dario
  • Joao, Jose A.
  • Moretó Planas, Miquel|||0000-0002-9848-8758
The irruption of chiplet-based architectures has been a game changer, enabling higher transistor integration and core counts in a single socket. However, chiplets impose higher and non-uniform memory access (NUMA) latencies than monolithic integration. This harms the efficiency of atomic memory operations (AMOs), which are fundamental to implementing fine-grained synchronization and concurrent data structures on large systems. AMOs are executed either near the core (near) or at a remote location within the cache hierarchy (far). On near AMOs, the core’s private cache fetches the target cache line in exclusiveness to modify it locally. Near AMOs cause significant data movement between private caches, especially harming parallel applications’ performance on chiplet-based architectures. Alternatively, far AMOs can alleviate the communication overhead by reducing data movement between processing elements. However, current multicore architectures only support one type of far AMO, which sends all updates to a single serialization point (centralized AMOs). This work introduces two new types of far AMOs, delegated and migrating, that execute AMOs remotely without centralizing updates in a single point of the cache hierarchy. Combining centralized, delegated, and migrating AMOs allows the directory to select the best location to execute AMOs. Moreover, we propose Delegato, a tracing optimization to effectively transport usage information from private caches to the directory to predict the best atomic type to issue accurately. Additionally, we design a simple predictor on top of Delegato that seamlessly selects the best placement to perform AMOs based on the data access pattern and usage activity of cores. Our evaluation using gem5 shows that Delegato can speed up applications on average by 1.07 × over centralized AMOs and by 1.13 × over the state-of-the-art AMO predictor., This research was supported by the Spanish Ministry of Science and Innovation through contracts [PID2023-146511NB-I00], and [PID2022-136454NB-C22] and by the Ministry for Digital Transformation and Public Service, within the framework of the Recovery, Transformation and Resilience Plan- NextGenerationEU [REGAGE22e00058408992]; the Generalitat of Catalunya through contract [2021-SGR-00763]; the Government of Aragón [T58_23R]; the Arm-BSC Center of Excellence, and the European Processor Initiative (EPI) which is part of the European Union’s Horizon 2020 research and innovation program under grant agreement No. 800928 and No. 101036168. V. Soria-Pardos has been supported through an FPU fellowship [FPU20-02132]; A. Armejach is a Serra Hunter Fellow., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

FetchFlare: An open-source strided data prefetcher for high-performance cache hierarchies

  • Korkian, Golnaz
  • Leyva Santes, Neiel Israel
  • Bigas Soldevilla, Arnau
  • Oliete Escuin, Noelia
  • Haghi, Abbas
  • Monemi, Alireza
  • Fuguet Tortolero, César
  • Alvarez Martí, Lluc
In recent years, the rise of open-source hardware has transformed the landscape of technology development. In particular, RISC-V has offered hardware designers the possibility of designing processors in a much cheaper way by leveraging a rich ecosystem of open-source designs that can be easily reused, extended, and customized. Although the RISC-V ecosystem is rapidly growing and open-source processors are becoming increasingly sophisticated, some advanced architectural techniques typically employed in commercial high-performance processors are still not prevalent in RISC-V open-source architectures. Among them, hardware prefetchers have been ubiquitous in highend processors for many years, but they are not as commonly found in open-source RISC-V processors. To bridge this gap, this work presents FetchFlare, a stride prefetcher for highperformance cache hierarchies. FetchFlare is able to capture the memory access patterns of applications, predict future memory accesses, and issue prefetch requests for them. We provide an open-source RTL implementation of FetchFlare and integrate it into a complete open-source setup formed by the OpenPiton framework, the Sargantana core, and the High-Performance Data Cache (HPDCache). Compared to a baseline system without prefetching, FetchFlare achieves an average speedup of 63%, avoids cache misses in the L1D and the L2 caches, and presents an average accuracy, coverage, and timeliness of 86%,39%, and 99%, respectively., This work has been partially supported by the European HiPEAC Network of Excellence, by the Generalitat de Catalunya (contract 2021-SGR-00763), and by Lenovo-BSC Contract-Framework Contract (2022). The BZL project is funded by the Ministerio de Transformación Digital y de la Función Pública by the Plan de Recuperación, Transformación y Resiliencia - financed by the European Union - NextGenerationEU. This work is part of the project PID2023-146511NB-I00 funded by the Spanish Ministry of Science, Innovation and Universities MCIU/AEI/10.13039/501100011033 and EU ERDF., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Squire: A general-purpose accelerator to exploit fine-grain parallelism on dependency-bound kernels

  • Langarita Benítez, Rubén|||0000-0002-6080-724X
  • Alastruey Benedé, Jesús
  • Ibáñez Marín, Pablo
  • Marco Sola, Santiago|||0000-0001-7951-3914
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Armejach Sanosa, Adrià|||0000-0003-2869-668X
Multiple HPC applications are often bottlenecked by compute-intensive kernels implementing complex dependency patterns (data-dependency bound). Traditional general-purpose accelerators struggle to effectively exploit fine-grain parallelism due to limitations in implementing convoluted data-dependency patterns (like SIMD) and overheads due to synchronization and data transfers (like GPGPUs). In contrast, custom FPGA and ASIC designs offer improved performance and energy efficiency at a high cost in hardware design and programming complexity and often lack the flexibility to process different workloads. We propose Squire, a general-purpose accelerator designed to exploit fine-grain parallelism effectively on dependency-bound kernels. Each Squire accelerator has a set of general-purpose low-power in-order cores that can rapidly communicate among themselves and directly access data from the L2 cache. Our proposal integrates one Squire accelerator per core in a typical multicore system, allowing the acceleration of dependency-bound kernels within parallel tasks with minimal software changes. As a case study, we evaluate Squire’s effectiveness by accelerating five kernels that implement complex dependency patterns. We use three of these kernels to build an end-to-end read-mapping tool that will be used to evaluate Squire. Squire obtains speedups up to 7.64× in dynamic programming kernels. Overall, Squire provides an acceleration for an end-to-end application of 3.66×. In addition, Squire reduces energy consumption by up to 56% with a minimal area overhead of 10.5% compared to a NeoverseN1 baseline., This research was supported by MICIU/AEI/10.13039/501100011033 and by ”ERDF A way of making Europe” through contracts [PID2023-146511NB-I00], [PID2023-146193OB-I00], and [PID2022-136454NBC22]; by the Ministry for Digital Transformation and Public Service, (i) via the framework of the Recovery, Transformation and Resilience Plan - NextGenerationEU [REGAGE22e00058408992] and (ii) through the Càtedra Chip UPC project [grant number TSI-069100-2023-0015]; by the Generalitat of Catalunya through contract [2021-SGR-00763]; by the Arm-BSC Center of Excellence; by the Lenovo-BSC Framework Contract (2020); and by Government of Aragón (T58 23R research group). A. Armejach is a Serra Hunter Fellow., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

SMX: Heterogeneous architecture for universal sequence alignment acceleration

  • Doblas Font, Max|||0000-0002-8909-3033
  • Shih, Pojui
  • Lostes Cazorla, Óscar|||0000-0001-5166-3806
  • Moretó Planas, Miquel|||0000-0002-9848-8758
  • Batten, Christopher
  • Marco Sola, Santiago|||0000-0001-7951-3914
Sequence alignment is a fundamental building block for critical applications across multiple fields, such as computational biology and information retrieval. The rapid advancement of genome sequencing technologies and breakthrough generative AI tools, like AlphaFold, has driven an exponential increase in sequence-data production, creating a pressing need for fast and efficient sequence alignment tools to analyze ever-growing biological sequence databases. Notwithstanding the numerous accelerators proposed, from general-purpose architectures (CPUs and GPUs) to domain-specific designs (FPGAs and ASICs), the most efficient solutions suffer from over-specialization and fail to adapt to the wide variety of irregular use cases demanded by practical sequence alignment applications. Thus, it remains a challenge to design an architecture that can balance efficiency and flexibility to meet the demands of real-world alignment applications. This work introduces SMX, a heterogeneous architecture designed for high-performance sequence alignment that supports various configurations for different sequence types (DNA, protein, and ASCII text) and alignment models (including weighted gaps and substitution matrices). SMX integrates an ISA extension (SMX-1D) for irregular and sequential tasks and a specialized coprocessor (SMX-2D) to accelerate regular and parallel tasks, both orchestrated by the general-purpose core to enable seamless integration with state-of-the-art sequence alignment algorithms. Our results demonstrate that SMX’s heterogeneous architecture accelerates different sequence alignment use cases by 256–744 × compared to state-of-the-art software implementations when aligning real datasets. Compared to specialized hardware accelerators, SMX delivers up to 18.5 × more peak performance per area added while providing greater flexibility to accelerate different use cases. Physical design results targeting a 22nm technology node estimate SMX’s area at 0.34mm2, which is only 30% of a single-issue in-order CPU. SMX offers a high-performance and efficient heterogeneous architecture for accelerating practical sequence alignment algorithms, providing a scalable and flexible solution tailored to meet the needs of modern sequence-analysis tools. Furthermore, an SMX case study explores the frontier between flexibility and efficiency in domain-specific architectures and accelerators., This work has been partially funded by the Spanish Ministry of Science and Innovation MCIN/AEI/10.13039/501100011033 and FSE+ [grant number PID2019-107255GB-C21, PID2020-113614RB-C21, TED2021-132634A-I00, and PID2023-146511NB-I00], the Spanish Ministry of Economy, Industry, and Competitiveness through an FPU fellowship [grant number FPU20-04076 to M.D.], LenovoBSC Contract-Framework Contract (2020), the NSF CSSI Award #2311890 and NSF PPoSS Award #2118709, the Spanish Ministry for Digital Transformation and Public Service, and the European Union – NextGenerationEU through the Càtedra Chip UPC project [grant number TSI-069100-2023-0015], the Spanish Ministry for Digital Transformation and Public Service, within the framework of the Recovery, Transformation and Resilience Plan - NextGenerationEU [REGAGE22e00058408992], and the Generalitat de Catalunya GenCat-DIUiE (GRR) [grant numbers 2021-SGR00763, and 2021-SGR-00574]., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

PQCUARK: A scalar RISC-V ISA extension for ML-KEM and ML-DSA

  • Carril Gil, Xavier|||0009-0008-8392-7579
  • Manuel Pasoot, Alicia
  • Parisi, Emanuele
  • Farràs Ventura, Oriol
  • Lara Niño, Carlos Andrés
  • Moretó Planas, Miquel|||0000-0002-9848-8758
© 2026 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works., Recent advances in quantum computing pose a threat to the security of digital communications, as large-scale quantum machines can break commonly used public-key cryptographic algorithms. To mitigate this risk, post-quantum cryptography (PQC) schemes are being standardized, with recent NIST recommendations selecting two lattice-based schemes, ML-KEM for key encapsulation and ML-DSA for digital signatures, alongside other schemes. Two computationally intensive kernels dominate the execution of these schemes: the Number-Theoretic Transform (NTT) for polynomial multiplication and the Keccak-f1600 permutation function for polynomial sampling and hashing. This paper presents PQCUARK, a scalar RISC-V ISA extension that accelerates these key operations. PQCUARK integrates two novel accelerators within the core pipeline: (i) a packed SIMD butterfly unit capable of performing NTT butterfly operations on 2×32bit or 4×16bit polynomial coefficients, and (ii) a permutation engine that delivers two Keccak rounds per cycle, hosting a private state and a direct interface to the core Load Store Unit, eliminating the need for a custom register file interface. We have integrated PQCUARK into an RV64 core and deployed it on an FPGA. Experimental results demonstrate that PQCUARK provides up to 10.1× speedup over the NIST baselines and 2.3× over the optimized software, and it outperforms similar state-of-the-art approaches between 1.4-12.3× in performance. ASIC synthesis in GF22-FDSOI technology shows a moderate core area increase of 8% at 1.2 GHz, with PQCUARK units being outside the critical path., Xavier Carril is supported by the predoctoral programme AGAUR-FI Joan Oró grant (2024 FI-1 00520), funded by the Generalitat de Catalunya (Department of Research and Universities) and the European Social Fund Plus. Emanuele Parisi is supported by AI4S fellowships from the “Generación D” initiative (Red.es, Ministerio para la Transformación Digital y de la Función Pública, C005/24-ED CV1), funded by EU NextGenerationEU funds through PRTR, partially funded by Generalitat de Catalunya [2021-SGR-00763], and by Spanish MCIU/AEI project PID2023-146511NB-I00 co-funded by EU ERDF. O. Farràs and C.A. Lara-Nino are supported by project HERMES funded by European Union NextGenerationEU/PRTR via INCIBE and the Spanish Ministry of Science and Innovation MCIN/AEI 10.13039/501100011033 (Grants ACITHEC PID2021-124928NB-I00, MATSE PID2024-156636NBC22). This work is part of the project PID2023-146511NB-I00 funded by the Spanish Ministry of Science, Innovation and Universities MCIU /AEI /10.13039/501100011033 and EU ERDF., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Floating-point atomic memory operations

  • Soria Pardos, Víctor
  • Armejach Sanosa, Adrià|||0000-0003-2869-668X
  • Suárez Gracía, Dario
  • Moretó Planas, Miquel|||0000-0002-9848-8758
Modern computing systems are increasingly adopting chiplet-based architectures, enabling the integration of different vendor devices, such as Central Processing Units (CPUs), Graphic Processing Units (GPUs), within a single package. These heterogeneous systems require efficient inter-chip synchronization mechanisms, including Atomic Memory Operations (AMOs). However, efficient synchronization demands that all integrated devices support the same AMO capabilities. While GPUs natively support floating-point AMOs (FAMOs), CPUs do not, creating an architectural asymmetry that leads to synchronization inefficiencies. To address this gap, Arm has recently introduced the Large System Floating-point Extension (LSFE), which introduces novel FAMO support to CPUs. This work characterizes and models the impact of FAMOs on graph processing, Machine Learning (ML), and High-Performance Computing (HPC) workloads. Our analysis reveals that a substantial fraction of AMOs target floating-point data and are currently implemented using inefficient compare-and-swap (CAS) built-ins. Replacing CAS-based constructs with hardware native FAMOs significantly improves performance across a wide range of workloads. Furthermore, we examine the trade-offs of executing FAMOs at different levels of the memory hierarchy, comparing execution in private caches (near) versus shared caches (far). Our results show that near FAMOs achieve an average speed-up of 1.3× speed-up, while far FAMOs deliver an average speed-up of 1.5×. Overall, we demonstrate that FAMOs can enable more efficient synchronization across key application domains., This work has been partially supported by the Spanish Ministry of Science, Innovation and Universities (PID2023-146511NB-I00), by MCIN/AEI/10.13039/501100011033 (grant PID2022-136454NB-C22), by Arm-BSC Center-of-Excelence (2025), by Generalitat de Catalunya (2021-SGR-00763), and by Government of Aragon (T58_23R research group). The Spanish Ministry of Economy, Industry and Competitiveness has partially supported V. Soria-Pardos through a FPU fellowship no. FPU20-02132., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

DRAM errors and cosmic rays: space invaders or science fiction?

  • Boixaderas Coderch, Isaac
  • Amaya, Jorge
  • Moré Codina, Sergi
  • Bartolomé Rodríguez, Javier
  • Vicente Dorca, David
  • Unsal, Osman Sabri
  • Gizopoulos, Dimitris
  • Carpenter, Paul Matthew
  • Radojkovic, Petar
  • Ayguadé Parra, Eduard|||0000-0002-5146-103X
It is widely accepted that cosmic rays are a plausible cause of DRAM errors in high-performance computing (HPC) systems, and various studies suggest that they could explain some aspects of the observed DRAM error behavior. However, this phenomenon is insufficiently studied in production environments. We analyze the correlations between cosmic rays and DRAM errors on two HPC clusters: a production supercomputer with server-class DDR3-1600 and a prototype with LPDDR3-1600 and no hardware error correction. Our error logs cover 2000 billion MB-hours for the MareNostrum 3 supercomputer and 135 million MB-hours for the Mont-Blanc prototype. Our analysis combines quantitative analysis, formal statistical methods and machine learning. We detect no indications that cosmic rays have any influence on the DRAM errors. To understand whether the findings are specific to systems under study, located at 100 meters above the sea level, the analysis should be repeated on other HPC clusters, especially the ones located on higher altitudes. Also, analysis can (and should) be applied to revisit and extend numerous previous studies which use cosmic rays as a hypothetical explanation for some aspects of the observed DRAM error behaviors., The research was supported by the Spanish Government through the contracts PID2019-107255GB-C21, PID2023-146511NB-I00 and CEX2021-001148-S, funded by MCIN/AEI/10.13039/501100011033. The work also received funding from the Department of Research and Universities of the Government of Catalonia (Code: 2021 SGR 00807, 2021 SGR 01007, 2021 SGR 01264). Paul Carpenter holds the Ramon y Cajal fellowship RYC2018-025628-I funded by MICIU/AEI/10.13039/501100011033 and ”ESF Investing in your future”. Support also came from the European Union’s Horizon Europe research and innovation programme under grant agreement No. 101097224 (REBECCA). Further financial assistance was provided by the Barcelona Zettascale Laboratory (reference REGAGE22e00058408992), which is financed by the Ministry for Digital Transformation and Public Services within the framework of the Resilience and Recovery Facility, and the European Union NextGenerationEU initiative. We acknowledge the NMDB database (www.nmdb.eu), founded under the European Union’s FP7 programme (contract no. 213007), for supplying part of the neutron monitor data., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Memory Sandbox 2.0: A framework for enabling HBM2e vs HBM2 performance and telemetry analysis on Xilinx FPGAs

  • Perdomo Hourné, Elias Augusto|||0000-0002-2279-6068
  • Teruel Jurado, Joan
  • Ahmed, Sajjad
  • Salami, Behzad
  • Cervero García, Teresa
  • Martorell Bofill, Xavier|||0000-0002-0417-3430
FPGA-based emulation provides a reconfigurable and efficient environment for evaluating SoC memory systems prior to full silicon deployment. Increasingly used to accelerate HPC and AI workloads, FPGAs enable hardware-level exploration of performance-critical subsystems such as memory, making them ideal platforms for prototyping accelerators in today’s heterogeneous computing landscape. Due to modern HPC and AI workloads’ increasing bandwidth and energy requirements, understanding and optimizing off-chip memory behavior, especially with High Bandwidth Memory (HBM), becomes critical. In this position paper, we present the ongoing development of Memory Sandbox 2.0, a modular, open-source framework for analyzing memory behavior across both performance (throughput, latency) and telemetry (temperature, power). Memory Sandbox 2.0 enhances the original tool with increased configurability, modular extensions for emerging technologies such as HBM2e, flexible support for varied access patterns, and integrated telemetry monitoring to guide energy-aware system design. We aim to support design space exploration of memory-access behavior and its architectural implications across platforms. While full experimental validation is still underway, we share early insights into challenges faced while profiling HBM2 and HBM2e on cutting-edge AMD Xilinx FPGAs such as the Versal V80 and Alveo U280. We also describe key architectural changes in recent HBM systems and how our tool models and exposes their behavior. Memory Sandbox 2.0 helps identify architectural bottlenecks and trade-offs early in the design cycle by enabling reproducible, insightful evaluation of memory-access patterns and telemetry data. This contributes to more effective and energy-efficient FPGA-based accelerator development, paving the way for more performant and energy-efficient HPC and AI solutions., This work has been co-financed by the Barcelona Zettascale Laboratory under project reference REGAGE22e00058408992, with support from the Spanish Ministry for Digital Transformation and Public Services, within the framework of the Recovery and Resilience Facility and the European Union– NextGenerationEU; from the research projects ST4HPC (PID2023-147979NB-C21) and the ACAP project (PID2023-146511NB-I00) supported by the Spanish Ministry of Science and Innovation (MICIU), the State Research Agency (AEI), and the European Regional Development Fund (FEDER). Additional support was provided by the Lenovo–BSC Contract Framework (2022) and the LOCA project (CEX2021-001148-S) funded by MCIN/AEI /10.13039/501100011033 and by “ERDF/EU”., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

WISEDRAM: a reliable bitwise in-DRAM accelerator

  • Soleimani, Mohammad Arman
  • Rohbani, Nezam
  • Cristal Kestelman, Adrián|||0000-0003-1277-9296
  • Unsal, Osman Sabri
  • Sarbazi-Azad, Hamid
Processing-in-Memory (PIM) aims to address the costly data movement between processing elements and memory subsystem, by computing simple operations inside DRAM in parallel. The large capacity, wide activation size during cell access, and the maturity of DRAM technology, make this technology a great choice for PIM techniques. Nonetheless, vulnerability to process variation and noises, internal leakage of the cells, and high latency in cell access, limit the utilization of processing in DRAMs for real-world applications. This work proposes a fast PIM technique, called WISEDRAM, which leverages one row of special cells, called X-cells, to enable in-DRAM bulk-bitwise operations. Unlike previous approaches, WISEDRAM retains the conventional DRAM cell access procedure, thereby ensuring the reliability of cell access for reads and writes at a level equivalent to that of conventional DRAMs. Compared with the state-of-the-art, WISEDRAM exhibits 22% reduction in average bitwise computation latency and a 71% improvement in XOR operation execution speed, while imposing an area overhead of 1.6%., This work is part of the project PID2023-146511NB-I00 funded by the Spanish Ministry of Science, Innovation and Universities MICIU /AEI /10.13039/501100011033 and EU ERDF. This publication is promoted by the Barcelona Zettascale Laboratory, backed by the Ministry for Digital Transformation and of Public Services, within the framework of the Recovery, Transformation, and Resilience Plan – funded by the European Union – NextGenerationEU., Peer Reviewed
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publicación Contribución de congreso (ConferenceOutput)

Squire: a general-purpose accelerator to exploit fine-grain parallelism on dependency-bound kernels

  • Langarita, Rubén
  • Alastruey-Benedé, Jesús
  • Ibáñez-Marín, Pablo
  • Marco-Sola, Santiago
  • Moretó, Miquel
  • Armejach, Adrià
Multiple HPC applications are often bottlenecked by compute-intensive kernels implementing complex dependency patterns (data-dependency bound). Traditional general-purpose accelerators struggle to effectively exploit fine-grain parallelism due to limitations in implementing convoluted data-dependency patterns (like SIMD) and overheads due to synchronization and data transfers (like GPGPUs). In contrast, custom FPGA and ASIC designs offer improved performance and energy efficiency at a high cost in hardware design and programming complexity and often lack the flexibility to process different workloads. We propose Squire, a general-purpose accelerator designed to exploit fine-grain parallelism effectively on dependency-bound kernels. Each Squire accelerator has a set of general-purpose low-power in-order cores that can rapidly communicate among themselves and directly access data from the L2 cache. Our proposal integrates one Squire accelerator per core in a typical multicore system, allowing the acceleration of dependency-bound kernels within parallel tasks with minimal software changes. As a case study, we evaluate Squire’s effectiveness by accelerating five kernels that implement complex dependency patterns. We use three of these kernels to build an end-to-end read-mapping tool that will be used to evaluate Squire. Squire obtains speedups up to 7.64× in dynamic programming kernels. Overall, Squire provides an acceleration for an end-to-end application of 3.66×. In addition, Squire reduces energy consumption by up to 56% with a minimal area overhead of 10.5% compared to a NeoverseN1 baseline.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publicación Artículo científico (JournalArticle)

Floating-point atomic memory operations

  • Soria-Pardos, Víctor
  • Armejach, Adrià
  • Suárez, Darío
  • Moretó, Miquel
Modern computing systems are increasingly adopting chiplet-based architectures, enabling the integration of different vendor devices, such as Central Processing Units (CPUs), Graphic Processing Units (GPUs), within a single package. These heterogeneous systems require efficient inter-chip synchronization mechanisms, including Atomic Memory Operations (AMOs). However, efficient synchronization demands that all integrated devices support the same AMO capabilities. While GPUs natively support floating-point AMOs (FAMOs), CPUs do not, creating an architectural asymmetry that leads to synchronization inefficiencies. To address this gap, Arm has recently introduced the Large System Floating-point Extension (LSFE), which introduces novel FAMO support to CPUs. This work characterizes and models the impact of FAMOs on graph processing, Machine Learning (ML), and High-Performance Computing (HPC) workloads. Our analysis reveals that a substantial fraction of AMOs target floating-point data and are currently implemented using inefficient compare-and-swap (CAS) built-ins. Replacing CAS-based constructs with hardware native FAMOs significantly improves performance across a wide range of workloads. Furthermore, we examine the trade-offs of executing FAMOs at different levels of the memory hierarchy, comparing execution in private caches (near) versus shared caches (far). Our results show that near FAMOs achieve an average speed-up of 1.3× speed-up, while far FAMOs deliver an average speed-up of 1.5×. Overall, we demonstrate that FAMOs can enable more efficient synchronization across key application domains.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publicación Artículo científico (JournalArticle)

Singletrack: an algorithm for improving memory consumption and performance of gap-affine sequence alignment

  • López-Villellas, Lorién
  • Iñiguez, Cristian
  • Jiménez-Blanco, Albert
  • Aguado-Puig, Quim
  • Moretó, Miquel
  • Alastruey-Benedé, Jesús
  • Ibáñez, Pablo
  • Marco-Sola, Santiago
Motivation Advances in DNA sequencing have outpaced advances in computation, making sequence alignment a major bottleneck in genome data analyses. Classical dynamic programming (DP) algorithms are particularly memory-intensive, especially when computing gap-affine and dual gap-affine alignments. Existing strategies to reduce memory consumption often sacrifice speed or alignment accuracy. Results We present Singletrack, an efficient algorithm for backtrace gap-affine and dual gap-affine alignments that requires storing a single DP matrix while preserving optimal alignment results. Compared to classical DP algorithms, Singletrack removes the need to store additional matrices (i.e. 2 for gap-affine and 4 for dual gap-affine), significantly reducing memory consumption and, in turn, reducing pressure on the memory hierarchy and improving overall performance. Most importantly, Singletrack is a general backtrace method compatible with state-of-the-art DP-based algorithms and heuristics, such as the Suzuki-Kasahara (SK) and the Wavefront Alignment (WFA) algorithms. We demonstrate that Singletrack reduces memory consumption for both SK and WFA algorithms, lowering SK usage by 2× and 4× and WFA usage by 3× and 5× for gap-affine and dual gap-affine alignments, respectively. Moreover, replacing KSW2’s memory-reduction technique with Singletrack accelerates its SK implementation by up to 1.4× at the cost of doubling memory consumption, while Singletrack increases the performance of the WFA implementation in WFA2-lib by 1.2–2.1×. Compared to the efficient linear-memory BiWFA algorithm, the Singletrack-accelerated version of WFA trades a practical increase in memory usage for up to 5.2× higher performance.
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publicación Artículo científico (JournalArticle)

Theseus: fast and optimal affine-gap sequence-to-graph alignment

  • Jiménez-Blanco, Albert
  • López-Villellas, Lorién
  • Moure, Juan Carlos
  • Moreto, Miquel
  • Marco-Sola, Santiago
Motivation Sequence-to-graph alignment is a central problem in bioinformatics, with applications in multiple sequence alignment (MSA) and pangenome analysis, among others. However, current algorithms for optimal affine-gap alignment impose high memory and computational requirements, limiting their scalability to aligning long sequences to complex graphs. Practical solutions partially address this problem using heuristic strategies that ultimately trade off optimality for speed. Results This work presents Theseus, a novel, fast, and optimal affine-gap sequence-to-graph alignment algorithm. Theseus leverages similarities between genomic sequences to accelerate the alignment computation and reduces the overall memory requirements without compromising optimality. To that end, Theseus processes only a subset of the dynamic programming cells, using a sparse-data strategy that enables efficient sequence-to-graph alignment. Moreover, our algorithm supports optimal affine-gap alignment on arbitrary directed graphs, including those with cycles. We evaluate Theseus on two key problems: MSA and pangenome read mapping. For MSA, we compare it against SPOA, abPOA, and POASTA. Theseus is 1.6× to 17.6× faster than POASTA, and 7.3× faster, on average, than SPOA, both optimal aligners. Compared with abPOA, Theseus ensures optimality and scales to the largest problems. For pangenome read mapping, we benchmark Theseus against the alignment stage of the mapping tool vg map, along with the alignment kernels of SPOA, abPOA, and POASTA. Theseus outperforms the other methods, showing a 1.9× to 16.9× speedup on short reads. Moreover, Theseus is 1.5× to 36.3× faster than vg when aligning against synthetic cyclic graphs.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Floating-point atomic memory operations

  • Soria Pardos, Víctor
  • Armejach Sanosa, Adrià
  • Suárez Gracía, Dario
  • Moretó Planas, Miquel
Modern computing systems are increasingly adopting chiplet-based architectures, enabling the integration of different vendor devices, such as Central Processing Units (CPUs), Graphic Processing Units (GPUs), within a single package. These heterogeneous systems require efficient inter-chip synchronization mechanisms, including Atomic Memory Operations (AMOs). However, efficient synchronization demands that all integrated devices support the same AMO capabilities. While GPUs natively support floating-point AMOs (FAMOs), CPUs do not, creating an architectural asymmetry that leads to synchronization inefficiencies. To address this gap, Arm has recently introduced the Large System Floating-point Extension (LSFE), which introduces novel FAMO support to CPUs. This work characterizes and models the impact of FAMOs on graph processing, Machine Learning (ML), and High-Performance Computing (HPC) workloads. Our analysis reveals that a substantial fraction of AMOs target floating-point data and are currently implemented using inefficient compare-and-swap (CAS) built-ins. Replacing CAS-based constructs with hardware native FAMOs significantly improves performance across a wide range of workloads. Furthermore, we examine the trade-offs of executing FAMOs at different levels of the memory hierarchy, comparing execution in private caches (near) versus shared caches (far). Our results show that near FAMOs achieve an average speed-up of 1.3× speed-up, while far FAMOs deliver an average speed-up of 1.5×. Overall, we demonstrate that FAMOs can enable more efficient synchronization across key application domains., This work has been partially supported by the Spanish Ministry of Science, Innovation and Universities (PID2023-146511NB-I00), by MCIN/AEI/10.13039/501100011033 (grant PID2022-136454NB-C22), by Arm-BSC Center-of-Excelence (2025), by Generalitat de Catalunya (2021-SGR-00763), and by Government of Aragon (T58_23R research group). The Spanish Ministry of Economy, Industry and Competitiveness has partially supported V. Soria-Pardos through a FPU fellowship no. FPU20-02132., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

FetchFlare: An open-source strided data prefetcher for high-performance cache hierarchies

  • Korkian, Golnaz
  • Leyva Santes, Neiel Israel
  • Bigas Soldevilla, Arnau
  • Oliete Escuin, Noelia
  • Haghi, Abbas
  • Monemi, Alireza
  • Fuguet Tortolero, César
  • Alvarez Martí, Lluc
In recent years, the rise of open-source hardware has transformed the landscape of technology development. In particular, RISC-V has offered hardware designers the possibility of designing processors in a much cheaper way by leveraging a rich ecosystem of open-source designs that can be easily reused, extended, and customized. Although the RISC-V ecosystem is rapidly growing and open-source processors are becoming increasingly sophisticated, some advanced architectural techniques typically employed in commercial high-performance processors are still not prevalent in RISC-V open-source architectures. Among them, hardware prefetchers have been ubiquitous in highend processors for many years, but they are not as commonly found in open-source RISC-V processors. To bridge this gap, this work presents FetchFlare, a stride prefetcher for highperformance cache hierarchies. FetchFlare is able to capture the memory access patterns of applications, predict future memory accesses, and issue prefetch requests for them. We provide an open-source RTL implementation of FetchFlare and integrate it into a complete open-source setup formed by the OpenPiton framework, the Sargantana core, and the High-Performance Data Cache (HPDCache). Compared to a baseline system without prefetching, FetchFlare achieves an average speedup of 63%, avoids cache misses in the L1D and the L2 caches, and presents an average accuracy, coverage, and timeliness of 86%,39%, and 99%, respectively., This work has been partially supported by the European HiPEAC Network of Excellence, by the Generalitat de Catalunya (contract 2021-SGR-00763), and by Lenovo-BSC Contract-Framework Contract (2022). The BZL project is funded by the Ministerio de Transformación Digital y de la Función Pública by the Plan de Recuperación, Transformación y Resiliencia - financed by the European Union - NextGenerationEU. This work is part of the project PID2023-146511NB-I00 funded by the Spanish Ministry of Science, Innovation and Universities MCIU/AEI/10.13039/501100011033 and EU ERDF., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

PQCUARK: A scalar RISC-V ISA extension for ML-KEM and ML-DSA

  • Carril Gil, Xavier
  • Manuel Pasoot, Alicia
  • Parisi, Emanuele
  • Farràs Ventura, Oriol
  • Lara Niño, Carlos Andrés
  • Moretó Planas, Miquel
© 2026 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works., Recent advances in quantum computing pose a threat to the security of digital communications, as large-scale quantum machines can break commonly used public-key cryptographic algorithms. To mitigate this risk, post-quantum cryptography (PQC) schemes are being standardized, with recent NIST recommendations selecting two lattice-based schemes, ML-KEM for key encapsulation and ML-DSA for digital signatures, alongside other schemes. Two computationally intensive kernels dominate the execution of these schemes: the Number-Theoretic Transform (NTT) for polynomial multiplication and the Keccak-f1600 permutation function for polynomial sampling and hashing. This paper presents PQCUARK, a scalar RISC-V ISA extension that accelerates these key operations. PQCUARK integrates two novel accelerators within the core pipeline: (i) a packed SIMD butterfly unit capable of performing NTT butterfly operations on 2×32bit or 4×16bit polynomial coefficients, and (ii) a permutation engine that delivers two Keccak rounds per cycle, hosting a private state and a direct interface to the core Load Store Unit, eliminating the need for a custom register file interface. We have integrated PQCUARK into an RV64 core and deployed it on an FPGA. Experimental results demonstrate that PQCUARK provides up to 10.1× speedup over the NIST baselines and 2.3× over the optimized software, and it outperforms similar state-of-the-art approaches between 1.4-12.3× in performance. ASIC synthesis in GF22-FDSOI technology shows a moderate core area increase of 8% at 1.2 GHz, with PQCUARK units being outside the critical path., Xavier Carril is supported by the predoctoral programme AGAUR-FI Joan Oró grant (2024 FI-1 00520), funded by the Generalitat de Catalunya (Department of Research and Universities) and the European Social Fund Plus. Emanuele Parisi is supported by AI4S fellowships from the “Generación D” initiative (Red.es, Ministerio para la Transformación Digital y de la Función Pública, C005/24-ED CV1), funded by EU NextGenerationEU funds through PRTR, partially funded by Generalitat de Catalunya [2021-SGR-00763], and by Spanish MCIU/AEI project PID2023-146511NB-I00 co-funded by EU ERDF. O. Farràs and C.A. Lara-Nino are supported by project HERMES funded by European Union NextGenerationEU/PRTR via INCIBE and the Spanish Ministry of Science and Innovation MCIN/AEI 10.13039/501100011033 (Grants ACITHEC PID2021-124928NB-I00, MATSE PID2024-156636NBC22). This work is part of the project PID2023-146511NB-I00 funded by the Spanish Ministry of Science, Innovation and Universities MCIU /AEI /10.13039/501100011033 and EU ERDF., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Delegato: Locality-aware atomic memory operations on chiplets

  • Soria Pardos, Víctor
  • Armejach Sanosa, Adrià
  • Mück, Tiago
  • Suárez Gracía, Dario
  • Joao, Jose A.
  • Moretó Planas, Miquel
The irruption of chiplet-based architectures has been a game changer, enabling higher transistor integration and core counts in a single socket. However, chiplets impose higher and non-uniform memory access (NUMA) latencies than monolithic integration. This harms the efficiency of atomic memory operations (AMOs), which are fundamental to implementing fine-grained synchronization and concurrent data structures on large systems. AMOs are executed either near the core (near) or at a remote location within the cache hierarchy (far). On near AMOs, the core’s private cache fetches the target cache line in exclusiveness to modify it locally. Near AMOs cause significant data movement between private caches, especially harming parallel applications’ performance on chiplet-based architectures. Alternatively, far AMOs can alleviate the communication overhead by reducing data movement between processing elements. However, current multicore architectures only support one type of far AMO, which sends all updates to a single serialization point (centralized AMOs). This work introduces two new types of far AMOs, delegated and migrating, that execute AMOs remotely without centralizing updates in a single point of the cache hierarchy. Combining centralized, delegated, and migrating AMOs allows the directory to select the best location to execute AMOs. Moreover, we propose Delegato, a tracing optimization to effectively transport usage information from private caches to the directory to predict the best atomic type to issue accurately. Additionally, we design a simple predictor on top of Delegato that seamlessly selects the best placement to perform AMOs based on the data access pattern and usage activity of cores. Our evaluation using gem5 shows that Delegato can speed up applications on average by 1.07 × over centralized AMOs and by 1.13 × over the state-of-the-art AMO predictor., This research was supported by the Spanish Ministry of Science and Innovation through contracts [PID2023-146511NB-I00], and [PID2022-136454NB-C22] and by the Ministry for Digital Transformation and Public Service, within the framework of the Recovery, Transformation and Resilience Plan- NextGenerationEU [REGAGE22e00058408992]; the Generalitat of Catalunya through contract [2021-SGR-00763]; the Government of Aragón [T58_23R]; the Arm-BSC Center of Excellence, and the European Processor Initiative (EPI) which is part of the European Union’s Horizon 2020 research and innovation program under grant agreement No. 800928 and No. 101036168. V. Soria-Pardos has been supported through an FPU fellowship [FPU20-02132]; A. Armejach is a Serra Hunter Fellow., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

DRAM errors and cosmic rays: space invaders or science fiction?

  • Boixaderas Coderch, Isaac
  • Amaya, Jorge
  • Moré Codina, Sergi
  • Bartolomé Rodríguez, Javier
  • Vicente Dorca, David
  • Unsal, Osman Sabri
  • Gizopoulos, Dimitris
  • Carpenter, Paul Matthew
  • Radojkovic, Petar
  • Ayguadé Parra, Eduard
It is widely accepted that cosmic rays are a plausible cause of DRAM errors in high-performance computing (HPC) systems, and various studies suggest that they could explain some aspects of the observed DRAM error behavior. However, this phenomenon is insufficiently studied in production environments. We analyze the correlations between cosmic rays and DRAM errors on two HPC clusters: a production supercomputer with server-class DDR3-1600 and a prototype with LPDDR3-1600 and no hardware error correction. Our error logs cover 2000 billion MB-hours for the MareNostrum 3 supercomputer and 135 million MB-hours for the Mont-Blanc prototype. Our analysis combines quantitative analysis, formal statistical methods and machine learning. We detect no indications that cosmic rays have any influence on the DRAM errors. To understand whether the findings are specific to systems under study, located at 100 meters above the sea level, the analysis should be repeated on other HPC clusters, especially the ones located on higher altitudes. Also, analysis can (and should) be applied to revisit and extend numerous previous studies which use cosmic rays as a hypothetical explanation for some aspects of the observed DRAM error behaviors., The research was supported by the Spanish Government through the contracts PID2019-107255GB-C21, PID2023-146511NB-I00 and CEX2021-001148-S, funded by MCIN/AEI/10.13039/501100011033. The work also received funding from the Department of Research and Universities of the Government of Catalonia (Code: 2021 SGR 00807, 2021 SGR 01007, 2021 SGR 01264). Paul Carpenter holds the Ramon y Cajal fellowship RYC2018-025628-I funded by MICIU/AEI/10.13039/501100011033 and ”ESF Investing in your future”. Support also came from the European Union’s Horizon Europe research and innovation programme under grant agreement No. 101097224 (REBECCA). Further financial assistance was provided by the Barcelona Zettascale Laboratory (reference REGAGE22e00058408992), which is financed by the Ministry for Digital Transformation and Public Services within the framework of the Resilience and Recovery Facility, and the European Union NextGenerationEU initiative. We acknowledge the NMDB database (www.nmdb.eu), founded under the European Union’s FP7 programme (contract no. 213007), for supplying part of the neutron monitor data., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Squire: A general-purpose accelerator to exploit fine-grain parallelism on dependency-bound kernels

  • Langarita Benítez, Rubén
  • Alastruey Benedé, Jesús
  • Ibáñez Marín, Pablo
  • Marco Sola, Santiago
  • Moretó Planas, Miquel
  • Armejach Sanosa, Adrià
Multiple HPC applications are often bottlenecked by compute-intensive kernels implementing complex dependency patterns (data-dependency bound). Traditional general-purpose accelerators struggle to effectively exploit fine-grain parallelism due to limitations in implementing convoluted data-dependency patterns (like SIMD) and overheads due to synchronization and data transfers (like GPGPUs). In contrast, custom FPGA and ASIC designs offer improved performance and energy efficiency at a high cost in hardware design and programming complexity and often lack the flexibility to process different workloads. We propose Squire, a general-purpose accelerator designed to exploit fine-grain parallelism effectively on dependency-bound kernels. Each Squire accelerator has a set of general-purpose low-power in-order cores that can rapidly communicate among themselves and directly access data from the L2 cache. Our proposal integrates one Squire accelerator per core in a typical multicore system, allowing the acceleration of dependency-bound kernels within parallel tasks with minimal software changes. As a case study, we evaluate Squire’s effectiveness by accelerating five kernels that implement complex dependency patterns. We use three of these kernels to build an end-to-end read-mapping tool that will be used to evaluate Squire. Squire obtains speedups up to 7.64× in dynamic programming kernels. Overall, Squire provides an acceleration for an end-to-end application of 3.66×. In addition, Squire reduces energy consumption by up to 56% with a minimal area overhead of 10.5% compared to a NeoverseN1 baseline., This research was supported by MICIU/AEI/10.13039/501100011033 and by ”ERDF A way of making Europe” through contracts [PID2023-146511NB-I00], [PID2023-146193OB-I00], and [PID2022-136454NBC22]; by the Ministry for Digital Transformation and Public Service, (i) via the framework of the Recovery, Transformation and Resilience Plan - NextGenerationEU [REGAGE22e00058408992] and (ii) through the Càtedra Chip UPC project [grant number TSI-069100-2023-0015]; by the Generalitat of Catalunya through contract [2021-SGR-00763]; by the Arm-BSC Center of Excellence; by the Lenovo-BSC Framework Contract (2020); and by Government of Aragón (T58 23R research group). A. Armejach is a Serra Hunter Fellow., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

WISEDRAM: a reliable bitwise in-DRAM accelerator

  • Soleimani, Mohammad Arman
  • Rohbani, Nezam
  • Cristal Kestelman, Adrián
  • Unsal, Osman Sabri
  • Sarbazi-Azad, Hamid
Processing-in-Memory (PIM) aims to address the costly data movement between processing elements and memory subsystem, by computing simple operations inside DRAM in parallel. The large capacity, wide activation size during cell access, and the maturity of DRAM technology, make this technology a great choice for PIM techniques. Nonetheless, vulnerability to process variation and noises, internal leakage of the cells, and high latency in cell access, limit the utilization of processing in DRAMs for real-world applications. This work proposes a fast PIM technique, called WISEDRAM, which leverages one row of special cells, called X-cells, to enable in-DRAM bulk-bitwise operations. Unlike previous approaches, WISEDRAM retains the conventional DRAM cell access procedure, thereby ensuring the reliability of cell access for reads and writes at a level equivalent to that of conventional DRAMs. Compared with the state-of-the-art, WISEDRAM exhibits 22% reduction in average bitwise computation latency and a 71% improvement in XOR operation execution speed, while imposing an area overhead of 1.6%., This work is part of the project PID2023-146511NB-I00 funded by the Spanish Ministry of Science, Innovation and Universities MICIU /AEI /10.13039/501100011033 and EU ERDF. This publication is promoted by the Barcelona Zettascale Laboratory, backed by the Ministry for Digital Transformation and of Public Services, within the framework of the Recovery, Transformation, and Resilience Plan – funded by the European Union – NextGenerationEU., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

SMX: Heterogeneous architecture for universal sequence alignment acceleration

  • Doblas Font, Max
  • Shih, Pojui
  • Lostes Cazorla, Óscar
  • Moretó Planas, Miquel
  • Batten, Christopher
  • Marco Sola, Santiago
Sequence alignment is a fundamental building block for critical applications across multiple fields, such as computational biology and information retrieval. The rapid advancement of genome sequencing technologies and breakthrough generative AI tools, like AlphaFold, has driven an exponential increase in sequence-data production, creating a pressing need for fast and efficient sequence alignment tools to analyze ever-growing biological sequence databases. Notwithstanding the numerous accelerators proposed, from general-purpose architectures (CPUs and GPUs) to domain-specific designs (FPGAs and ASICs), the most efficient solutions suffer from over-specialization and fail to adapt to the wide variety of irregular use cases demanded by practical sequence alignment applications. Thus, it remains a challenge to design an architecture that can balance efficiency and flexibility to meet the demands of real-world alignment applications. This work introduces SMX, a heterogeneous architecture designed for high-performance sequence alignment that supports various configurations for different sequence types (DNA, protein, and ASCII text) and alignment models (including weighted gaps and substitution matrices). SMX integrates an ISA extension (SMX-1D) for irregular and sequential tasks and a specialized coprocessor (SMX-2D) to accelerate regular and parallel tasks, both orchestrated by the general-purpose core to enable seamless integration with state-of-the-art sequence alignment algorithms. Our results demonstrate that SMX’s heterogeneous architecture accelerates different sequence alignment use cases by 256–744 × compared to state-of-the-art software implementations when aligning real datasets. Compared to specialized hardware accelerators, SMX delivers up to 18.5 × more peak performance per area added while providing greater flexibility to accelerate different use cases. Physical design results targeting a 22nm technology node estimate SMX’s area at 0.34mm2, which is only 30% of a single-issue in-order CPU. SMX offers a high-performance and efficient heterogeneous architecture for accelerating practical sequence alignment algorithms, providing a scalable and flexible solution tailored to meet the needs of modern sequence-analysis tools. Furthermore, an SMX case study explores the frontier between flexibility and efficiency in domain-specific architectures and accelerators., This work has been partially funded by the Spanish Ministry of Science and Innovation MCIN/AEI/10.13039/501100011033 and FSE+ [grant number PID2019-107255GB-C21, PID2020-113614RB-C21, TED2021-132634A-I00, and PID2023-146511NB-I00], the Spanish Ministry of Economy, Industry, and Competitiveness through an FPU fellowship [grant number FPU20-04076 to M.D.], LenovoBSC Contract-Framework Contract (2020), the NSF CSSI Award #2311890 and NSF PPoSS Award #2118709, the Spanish Ministry for Digital Transformation and Public Service, and the European Union – NextGenerationEU through the Càtedra Chip UPC project [grant number TSI-069100-2023-0015], the Spanish Ministry for Digital Transformation and Public Service, within the framework of the Recovery, Transformation and Resilience Plan - NextGenerationEU [REGAGE22e00058408992], and the Generalitat de Catalunya GenCat-DIUiE (GRR) [grant numbers 2021-SGR00763, and 2021-SGR-00574]., Peer Reviewed, Postprint (author's final draft)