Towards Efficient Integration of Auto-Generated Hardware Accelerators Models with Architectural Simulators
NCSU researchers have been developing an automated code generation system to generate models for accelerators that are suitable for simulation with the SST/macro simulator.
This project seeks to identify methods to integrate auto-generated accelerator models with the SST/macro simulator. NC State researchers have developed an automated code generation system to generate models for accelerators that are compatible with the SST/macro simulator. The goal is to enable a smoother transition from model generation to simulation for the user.
Sponsor
National Technology and Engineering Solutions of Sandia, LLC. (Sandia National Laboratories)
The grant—running from November 21, 2022 to September 30, 2023—is for a total of $100,000.
Principle Investigators
Amro Awad
More Details
he project aims to identify novel ways to seamlessly integrate auto-generated accelerator models with the SST simulator.