It is now possible to create an ESP instance with one or multiple accelerator tiles hosting the NVIDIA Deep Learning Accelerator NVDLA.
This is the first example showing how ESP can integrate third-party accelerators in any accelerator tile thanks to a new accelerator socket based on the open standard AXI bus protocol.
The accelerator device driver and application are integrated as well into the ESP flow and run on the RISC-V processor Ariane.