Product Details
autoAx: An Open-Source Automated Design Space Exploration Framework for Approximate Accelerators in FPGAs and ASICs
Created: 2023
Czech title
autoAx: Automatizovaný framework pro automatické prohledávání návrhového prostoru pro aproximativní akcelerátory v obvodech FPGA a ASIC s otevřeným zdrojovým kódem
Type
software
License
no - free
Authors
Mrázek Vojtěch, Ing., Ph.D. (DCSY FIT BUT)
Keywords
approximate computing, high level synthesis, machine learning
Description
The automated generation of approximate circuits and accelerators has been a useful design strategy to achieve energy efficiency and/or performance improvements. In this work, we propose a framework, autoAx, that leverages machine learning models that evaluate the state-of-the-art approximate components to explore the architecture space effectively. These accelerators are modeled at RTL and optimized using an evolutionary algorithm. The AutoAx framework is extensible, open-source, and can assist in exploring new directions in high-level approximation.
Location
Licence
Projects
Application-specific HW/SW architectures and their applications (FIT-S-23-8141)
Research groups
Evolvable Hardware Research Group (VZ EHW)
Departments
Department of Computer Systems FIT BUT (DCSY FIT BUT)