Hybrid Precision Floating-Point (HPFP) Selection to Optimize Hardware-Constrained Accelerator for CNN Training

Hybrid Precision Floating-Point (HPFP) Selection to Optimize Hardware-Constrained Accelerator for CNN Training

The rapid advancement in AI requires efficient accelerators for training on edge devices, which often face challenges related to the high hardware costs of floating-point arithmetic operations. To tackle these problems, efficient floating-point formats inspired by block floating-point (BFP), such as...

Full description

Bibliographic Details
Main Authors: Muhammad Junaid, Hayotjon Aliev, SangBo Park, HyungWon Kim, Hoyoung Yoo, Sanghoon Sim
Format: Article
Language:English
Published: MDPI AG 2024-03-01
Series:Sensors
Subjects:
floating points
BFP
MSFP
HPFP
deep neural network (DNN)
Yolov2
Online Access:https://www.mdpi.com/1424-8220/24/7/2145

Internet

https://www.mdpi.com/1424-8220/24/7/2145

Similar Items