带有倾斜流水线的脉动阵列中的低精度浮点算术
Reduced-Precision Floating-Point Arithmetic in Systolic Arrays with Skewed Pipelines
AI总结:
本文针对脉动阵列中低精度浮点链式乘加运算,提出一种倾斜流水线设计,通过为指数逻辑启用新前推路径实现相邻处理单元流水线阶段的并行执行,在极小硬件开销下显著降低矩阵乘法延迟,并在最先进CNN上实现8%至11%的能耗降低。
摘要:
硬件中深度学习内核的加速依赖于在脉动阵列(SA)上高效执行的矩阵乘法。为了有效地权衡深度学习训练/推理质量与硬件成本,SA加速器采用了低精度浮点(FP)算术。在本研究中,我们展示了对新流水线组织的需求,以降低延迟并提高由SA结构所强加的链式乘加运算的低精度FP算子的能效。所提出的倾斜流水线设计重新组织了FP乘加单元的流水线操作,为指数逻辑启用了新的前推路径,从而允许连续PE的流水线阶段并行执行。结果,SA内矩阵乘法运算的延迟以极小的硬件成本显著降低,从而在所检查的最先进CNN中实现了8%和11%的能耗降低。
英文摘要:
The acceleration of deep-learning kernels in hardware relies on matrix multiplications that are executed efficiently on Systolic Arrays (SA). To effectively trade off deep-learning training/inference quality with hardware cost, SA accelerators employ reduced-precision Floating-Point (FP) arithmetic. In this work, we demonstrate the need for new pipeline organizations to reduce latency and improve energy efficiency of reduced-precision FP operators for the chained multiply-add operation imposed by the structure of the SA. The proposed skewed pipeline design reorganizes the pipelined operation of the FP multiply-add units to enable new forwarding paths for the exponent logic, which allow for parallel execution of the pipeline stages of consecutive PEs. As a result, the latency of the matrix multiplication operation within the SA is significantly reduced with minimal hardware cost, thereby yielding an energy reduction of 8% and 11% for the examined state-of-the-art CNNs.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。