Logo image
LV: Latency-Versatile Floating-Point Engine for High-Performance Deep Neural Networks
期刊文章   同儕審查

LV: Latency-Versatile Floating-Point Engine for High-Performance Deep Neural Networks

Yun-Chen Lo, Yu-Chih TsaiRen-Shuo Liu
IEEE Computer Architecture Letters
2023

摘要

Adders Approximate computation Artificial neural networks Clocks Computer architecture Electric breakdown Engines floating point latency-versatile architecture Registers Hardware and Architecture
Computing latency is an important system metric for Deep Neural Networks (DNNs) accelerators. To reduce latency, this work proposes LV, a latency-versatile floating-point engine (FP-PE), which contains the following key contributions: 1) an approximate bit-versatile multiplier-and-accumulate (BV-MAC) unit with early shifter and 2) an on-demand fixed-point-to-floating-point conversion (FXP2FP) unit. The extensive experimental results show that LV outperforms baseline FP-PE and redundancy-aware FP-PE by up to 2.12&null and 1.3&null speedup using TSMC 40-nm technology, achieving comparable accuracy on the ImageNet classification tasks.

相關連結

指標

1 檢視次數

詳細資料

Logo image