AraXL: A Physically Scalable, Ultra-Wide RISC-V Vector Processor Design for Fast and Efficient Computation on Long Vectors

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Purayil, Navaneeth Kunhi, Perotti, Matteo, Fischer, Tim, Benini, Luca
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866912534924099584
author Purayil, Navaneeth Kunhi
Perotti, Matteo
Fischer, Tim
Benini, Luca
author_facet Purayil, Navaneeth Kunhi
Perotti, Matteo
Fischer, Tim
Benini, Luca
contents The ever-growing scale of data parallelism in today's HPC and ML applications presents a big challenge for computing architectures' energy efficiency and performance. Vector processors address the scale-up challenge by decoupling Vector Register File (VRF) and datapath widths, allowing the VRF to host long vectors and increase register-stored data reuse while reducing the relative cost of instruction fetch and decode. However, even the largest vector processor designs today struggle to scale to more than 8 vector lanes with double-precision Floating Point Units (FPUs) and 256 64-bit elements per vector register. This limitation is induced by difficulties in the physical implementation, which becomes wire-dominated and inefficient. In this work, we present AraXL, a modular and scalable 64-bit RISC-V V vector architecture targeting long-vector applications for HPC and ML. AraXL addresses the physical scalability challenges of state-of-the-art vector processors with a distributed and hierarchical interconnect, supporting up to 64 parallel vector lanes and reaching the maximum Vector Register File size of 64 Kibit/vreg permitted by the RISC-V V 1.0 ISA specification. Implemented in a 22-nm technology node, our 64-lane AraXL achieves a performance peak of 146 GFLOPs on computation-intensive HPC/ML kernels (>99% FPU utilization) and energy efficiency of 40.1 GFLOPs/W (1.15 GHz, TT, 0.8V), with only 3.8x the area of a 16-lane instance.
format Preprint
id arxiv_https___arxiv_org_abs_2501_10301
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle AraXL: A Physically Scalable, Ultra-Wide RISC-V Vector Processor Design for Fast and Efficient Computation on Long Vectors
Purayil, Navaneeth Kunhi
Perotti, Matteo
Fischer, Tim
Benini, Luca
Hardware Architecture
The ever-growing scale of data parallelism in today's HPC and ML applications presents a big challenge for computing architectures' energy efficiency and performance. Vector processors address the scale-up challenge by decoupling Vector Register File (VRF) and datapath widths, allowing the VRF to host long vectors and increase register-stored data reuse while reducing the relative cost of instruction fetch and decode. However, even the largest vector processor designs today struggle to scale to more than 8 vector lanes with double-precision Floating Point Units (FPUs) and 256 64-bit elements per vector register. This limitation is induced by difficulties in the physical implementation, which becomes wire-dominated and inefficient. In this work, we present AraXL, a modular and scalable 64-bit RISC-V V vector architecture targeting long-vector applications for HPC and ML. AraXL addresses the physical scalability challenges of state-of-the-art vector processors with a distributed and hierarchical interconnect, supporting up to 64 parallel vector lanes and reaching the maximum Vector Register File size of 64 Kibit/vreg permitted by the RISC-V V 1.0 ISA specification. Implemented in a 22-nm technology node, our 64-lane AraXL achieves a performance peak of 146 GFLOPs on computation-intensive HPC/ML kernels (>99% FPU utilization) and energy efficiency of 40.1 GFLOPs/W (1.15 GHz, TT, 0.8V), with only 3.8x the area of a 16-lane instance.
title AraXL: A Physically Scalable, Ultra-Wide RISC-V Vector Processor Design for Fast and Efficient Computation on Long Vectors
topic Hardware Architecture
url https://arxiv.org/abs/2501.10301