Exploring the Performance Improvement of Tensor Processing Engines through Transformation in the Bit-weight Dimension of MACs

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wu, Qizhe, Liang, Huawen, Gui, Yuchen, Zeng, Zhichen, He, Zerong, Tao, Linfeng, Wang, Xiaotian, Zhao, Letian, Zeng, Zhaoxi, Yuan, Wei, Wu, Wei, Jin, Xi
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909531839135744
author Wu, Qizhe
Liang, Huawen
Gui, Yuchen
Zeng, Zhichen
He, Zerong
Tao, Linfeng
Wang, Xiaotian
Zhao, Letian
Zeng, Zhaoxi
Yuan, Wei
Wu, Wei
Jin, Xi
author_facet Wu, Qizhe
Liang, Huawen
Gui, Yuchen
Zeng, Zhichen
He, Zerong
Tao, Linfeng
Wang, Xiaotian
Zhao, Letian
Zeng, Zhaoxi
Yuan, Wei
Wu, Wei
Jin, Xi
contents General matrix-matrix multiplication (GEMM) is a cornerstone of AI computations, making tensor processing engines (TPEs) increasingly critical in GPUs and domain-specific architectures. Existing architectures primarily optimize dataflow or operand reuse strategies. However, considering the interaction between matrix multiplication and multiply-accumulators (MACs) offers greater optimization potential. This work introduces a novel hardware perspective on matrix multiplication, focusing on the bit-weight dimension of MACs. We propose a finer-grained TPE notation using matrix triple loops as an example, introducing new methods for designing and optimizing PE microarchitectures. Based on this notation and its transformations, we propose four optimization techniques that improve timing, area, and power consumption. Implementing our design in RTL using the SMIC-28nm process, we evaluate its effectiveness across four classic TPE architectures: systolic array, 3D-Cube, multiplier-adder tree, and 2D-Matrix. Our techniques achieve area efficiency improvements of 1.27x, 1.28x, 1.56x, and 1.44x, and energy efficiency gains of 1.04x, 1.56x, 1.49x, and 1.20x, respectively. Applied to a bit-slice architecture, our approach achieves a 12.10x improvement in energy efficiency and 2.85x in area efficiency compared to Laconic. Our Verilog HDL code, along with timing, area, and power reports, is available at https://github.com/wqzustc/High-Performance-Tensor-Processing-Engines
format Preprint
id arxiv_https___arxiv_org_abs_2503_06342
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Exploring the Performance Improvement of Tensor Processing Engines through Transformation in the Bit-weight Dimension of MACs
Wu, Qizhe
Liang, Huawen
Gui, Yuchen
Zeng, Zhichen
He, Zerong
Tao, Linfeng
Wang, Xiaotian
Zhao, Letian
Zeng, Zhaoxi
Yuan, Wei
Wu, Wei
Jin, Xi
Hardware Architecture
Machine Learning
General matrix-matrix multiplication (GEMM) is a cornerstone of AI computations, making tensor processing engines (TPEs) increasingly critical in GPUs and domain-specific architectures. Existing architectures primarily optimize dataflow or operand reuse strategies. However, considering the interaction between matrix multiplication and multiply-accumulators (MACs) offers greater optimization potential. This work introduces a novel hardware perspective on matrix multiplication, focusing on the bit-weight dimension of MACs. We propose a finer-grained TPE notation using matrix triple loops as an example, introducing new methods for designing and optimizing PE microarchitectures. Based on this notation and its transformations, we propose four optimization techniques that improve timing, area, and power consumption. Implementing our design in RTL using the SMIC-28nm process, we evaluate its effectiveness across four classic TPE architectures: systolic array, 3D-Cube, multiplier-adder tree, and 2D-Matrix. Our techniques achieve area efficiency improvements of 1.27x, 1.28x, 1.56x, and 1.44x, and energy efficiency gains of 1.04x, 1.56x, 1.49x, and 1.20x, respectively. Applied to a bit-slice architecture, our approach achieves a 12.10x improvement in energy efficiency and 2.85x in area efficiency compared to Laconic. Our Verilog HDL code, along with timing, area, and power reports, is available at https://github.com/wqzustc/High-Performance-Tensor-Processing-Engines
title Exploring the Performance Improvement of Tensor Processing Engines through Transformation in the Bit-weight Dimension of MACs
topic Hardware Architecture
Machine Learning
url https://arxiv.org/abs/2503.06342