Tool Release: Argus – An End-to-End Framework for Accelerating CNNs on FPGAs
We are pleased to announce the release of the Argus online Verilog Code Generator tool. Argus is an end-to-end framework for accelerating CNNs on FPGAs. The core of Argus is an accelerator generator that translates high-level CNN descriptions into efficient multi-core accelerator designs. Argus explores an extensive design space, jointly optimizing all design aspects for the target FPGA and generating multi-core accelerator designs that achieve near-perfect dynamic arithmetic unit utilization.
Our online code generator allows anyone to use Argus to produce network-optimized and FPGA-optimized CNN accelerators in Verilog. To minimize user effort, Argus includes a model parser for importing CNN models from popular machine learning frameworks and a software stack for running an FPGA-backed CNN inference microservice. The tool can be accessed online at https://argus-webgen.compas.cs.stonybrook.edu/
The Argus tool was built by researchers in the COMPAS Lab at Stony Brook University’s departments of Computer Science and Electrical and Computer Engineering. Argus is part of a larger research effort studying efficient hardware acceleration of machine learning.