EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Mentor: A Memory-Efficient Sparse-dense Matrix Multiplication Accelerator Based on Column-Wise Product

Xiaobo Lu, Jianbin Fang, Lin Peng, Chun Huang, Zidong Du, Yongwei Zhao, Zheng Wang

ACM Transactions on Architecture and Code Optimization · 31.12.2024 · 21 (4) · 1-25

DOI: 10.1145/3688612 ↗

Kurz gefasst

Fachartikel in der Zeitschrift ACM Transactions on Architecture and Code Optimization, online erschienen am 20.11.2024. Verfasst von Xiaobo Lu und 6 weiteren Autor:innen. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 63 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Key Research and Development Program of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Sparse-dense matrix multiplication (SpMM) is the performance bottleneck of many high-performance and deep-learning applications, making it attractive to design specialized SpMM hardware accelerators. Unfortunately, existing hardware solutions do not take full advantage of data reuse opportunities of the input and output matrices or suffer from irregular memory access patterns. Their strategies increase the off-chip memory traffic and bandwidth pressure, leaving much room for improvement. We present Mentor , a new approach to designing SpMM accelerators. Our key insight is that column-wise dataflow, while rarely exploited in prior works, can address these issues in SpMM computations. Mentor is a software-hardware co-design approach for leveraging column-wise dataflow to improve data reuse and regular memory accesses of SpMM. On the software level, Mentor incorporates a novel streaming construction scheme to preprocess the input matrix for enabling a streaming access pattern. On the hardware level, it employs a fully pipelined design to unlock the potential of column-wise dataflow further. The design of Mentor is underpinned by a carefully designed analytical model to find the tradeoff between performance and hardware resources. We have implemented an FPGA prototype of Mentor . Experimental results show that Mentor achieves speedup by geomean 2.05× (up to 3.98×), reduces the memory traffic by geomean 2.92× (up to 4.93×), and improves bandwidth utilization by geomean 1.38× (up to 2.89×), compared with the state-of-the-art hardware solutions.

Förderung

  • National Key Research and Development Program of China

Zitieren

APA

Lu, X., Fang, J., Peng, L., Huang, C., Du, Z., Zhao, Y., & Wang, Z. (2024). Mentor: A Memory-Efficient Sparse-dense Matrix Multiplication Accelerator Based on Column-Wise Product. ACM Transactions on Architecture and Code Optimization, 21(4), 1-25. https://doi.org/10.1145/3688612

Vancouver

Lu X, Fang J, Peng L, Huang C, Du Z, Zhao Y, et al. Mentor: A Memory-Efficient Sparse-dense Matrix Multiplication Accelerator Based on Column-Wise Product. ACM Transactions on Architecture and Code Optimization. 2024;21(4):1-25. doi:10.1145/3688612

BibTeX
@article{Lu2024,
  author = {Lu, Xiaobo and Fang, Jianbin and Peng, Lin and Huang, Chun and Du, Zidong and Zhao, Yongwei and Wang, Zheng},
  title = {Mentor: A Memory-Efficient Sparse-dense Matrix Multiplication Accelerator Based on Column-Wise Product},
  journal = {ACM Transactions on Architecture and Code Optimization},
  year = {2024},
  volume = {21},
  number = {4},
  pages = {1--25},
  doi = {10.1145/3688612}
}
Kostenlos starten →