This wraps up the NVIDIA Deep Learning Accelerator (NVDLA) (https://github.com/nvdla) into a Rocket Chip based peripheral to be used in Chipyard.
This is based of the work done by the Computer Systems Design Laboratory at the University of Kansas (https://github.com/CSL-KU/nvidia-dla-blocks) and the original SiFive NVDLA wrapper repository (https://github.com/sifive/block-nvdla-sifive).
Please visit firesim-collateral for more information on how to build an FPGA image with NVDLA.