This technical paper titled “Gemmini: Enabling Systematic Deep-Learning Architecture Evaluation via Full-Stack Integration” was published jointly by researchers at UC Berkeley and a co-author from MIT. The research was partially funded by DARPA and won DAC 2021 Best Paper.
The paper presents Gemmini, “an open-source, full-stack DNN accelerator generator for DNN workloads, enabling end-to-end, fullstack implementation and evaluation of custom hardware accelerator systems for rapidly evolving DNN workloads. Gemmini’s hardware template and parameterization allows users to tune the hardware design options across a broad spectrum spanning performance, efficiency, and extensibility.”
Find the technical paper here and the DAC presentation video here.. Gemmini is open-sourced here.
Visit Semiconductor Engineering’s Technical Paper library here and discover many more chip industry academic papers.
Leave a Reply