Multiple Constant Multiplication on Neural Networks
Dnr:
NAISS 2024/22-158
Type:
NAISS Small Compute
Principal Investigator:
Xiaoyue Chen
Affiliation:
Uppsala universitet
Start Date:
2024-01-30
End Date:
2025-02-01
Primary Classification:
10206: Computer Engineering
Webpage:
Allocation
Abstract
This project investigates how Multiple Constant Multiplication (MCM) could be used to accelerate the computations of Neural Networks. The compute resource will be used for MCM optimization solvers. GPUs will also be needed for train and evaluate the inference of benchmark models.