Hardware vs. Software Implementation of Warp-Level Features in Vortex RISC-V GPU

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Pu, Huanzhi, Ravi, Rishabh, Jeong, Shinnung, Subramanya, Udit, Chung, Euijun, Zhao, Jisheng, Ahn, Chihyo, Kim, Hyesoon
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912362014965760
author Pu, Huanzhi
Ravi, Rishabh
Jeong, Shinnung
Subramanya, Udit
Chung, Euijun
Zhao, Jisheng
Ahn, Chihyo
Kim, Hyesoon
author_facet Pu, Huanzhi
Ravi, Rishabh
Jeong, Shinnung
Subramanya, Udit
Chung, Euijun
Zhao, Jisheng
Ahn, Chihyo
Kim, Hyesoon
contents RISC-V GPUs present a promising path for supporting GPU applications. Traditionally, GPUs achieve high efficiency through the SPMD (Single Program Multiple Data) programming model. However, modern GPU programming increasingly relies on warp-level features, which diverge from the conventional SPMD paradigm. In this paper, we explore how RISC-V GPUs can support these warp-level features both through hardware implementation and via software-only approaches. Our evaluation shows that a hardware implementation achieves up to 4 times geomean IPC speedup in microbenchmarks, while software-based solutions provide a viable alternative for area-constrained scenarios.
format Preprint
id arxiv_https___arxiv_org_abs_2505_03102
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Hardware vs. Software Implementation of Warp-Level Features in Vortex RISC-V GPU
Pu, Huanzhi
Ravi, Rishabh
Jeong, Shinnung
Subramanya, Udit
Chung, Euijun
Zhao, Jisheng
Ahn, Chihyo
Kim, Hyesoon
Hardware Architecture
RISC-V GPUs present a promising path for supporting GPU applications. Traditionally, GPUs achieve high efficiency through the SPMD (Single Program Multiple Data) programming model. However, modern GPU programming increasingly relies on warp-level features, which diverge from the conventional SPMD paradigm. In this paper, we explore how RISC-V GPUs can support these warp-level features both through hardware implementation and via software-only approaches. Our evaluation shows that a hardware implementation achieves up to 4 times geomean IPC speedup in microbenchmarks, while software-based solutions provide a viable alternative for area-constrained scenarios.
title Hardware vs. Software Implementation of Warp-Level Features in Vortex RISC-V GPU
topic Hardware Architecture
url https://arxiv.org/abs/2505.03102