EUVIMEDEUVIMED entdecken
Crossref

HighP: In-Memory Acceleration of SpGEMM With High Bank-Level Parallelism

Dan Chen, Huize Li, Huiying Lan, Zhaoying Li, Pengcheng Yao, Tulika Mitra

IEEE Transactions on Computers · 2026-01 · 75 (1) · 139-151

DOI: 10.1109/tc.2025.3624920 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Computers, 2026-01. Verfasst von Dan Chen und 5 weiteren Autor:innen. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 40 Literaturangaben. Gefördert von: National Research Foundation Singapore, Competitive Research Program, Ministry of Education - Singapore.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Generalized sparse matrix-matrix multiplication (SpGEMM) is a critical computational primitive that is highly memory-bound due to its inherent irregular data-dependent access pattern. Near-bank processing-in-memory (PIM) is a promising technique to overcome the memory bottleneck of SpGEMM by performing computations near the bank where the data is stored. However, earlier PIM studies fail to fully utilize the high memory bandwidth when performing SpGEMM due to low bank-level parallelism. As a result, 80% memory bandwidth is wasted as observed in our in-depth experimental analysis. Our key insight in this paper is that non-conflicting matrix columns in SpGEMM, where each row of these columns has no more than one non-zero element, can be processed simultaneously in different banks. We hence propose HighP, a near-bank PIM accelerator for SpGEMM with high bank-level parallelism. We first propose a set-based search mechanism, which finds non-conflicting columns through set operations automatically. We then develop a DIMM-based PIM architecture with detailed hardware and workflow designs for SpGEMM. Set operation logic and unified scratchpad memory management are designed to perform set operations with high computational parallelism and to enhance data reuse, respectively. HighP provides up to 17.88 $\times$ × performance improvement compared to the state-of-the-art SpGEMM accelerator and achieves up to 8.19 $\times$ × performance improvement over the state-of-the-art PIM solution.

Förderung

  • National Research Foundation Singapore
  • Competitive Research Program
  • Ministry of Education - Singapore

Zitieren

APA

Chen, D., Li, H., Lan, H., Li, Z., Yao, P., & Mitra, T. (2026). HighP: In-Memory Acceleration of SpGEMM With High Bank-Level Parallelism. IEEE Transactions on Computers, 75(1), 139-151. https://doi.org/10.1109/tc.2025.3624920

Vancouver

Chen D, Li H, Lan H, Li Z, Yao P, Mitra T. HighP: In-Memory Acceleration of SpGEMM With High Bank-Level Parallelism. IEEE Transactions on Computers. 2026;75(1):139-151. doi:10.1109/tc.2025.3624920

BibTeX
@article{Chen2026,
  author = {Chen, Dan and Li, Huize and Lan, Huiying and Li, Zhaoying and Yao, Pengcheng and Mitra, Tulika},
  title = {HighP: In-Memory Acceleration of SpGEMM With High Bank-Level Parallelism},
  journal = {IEEE Transactions on Computers},
  year = {2026},
  volume = {75},
  number = {1},
  pages = {139--151},
  doi = {10.1109/tc.2025.3624920}
}
Kostenlos starten →