How to extend NNVM/TVM as a static compiler for HW accelerator like NVDLA?

Questions
6

@thierry was there any progress or task for NVDLA+TVM integration?