Startup OctoML says that optimizing the compute graph of neural nets for new hardware may change even the training of neural nets.
In the latest development in the hardware and research feedback loop, Apache TVM’s process of optimization may already be shaping aspects of how AI is developed.
OctoML is in the field of MLOps, helping to operationalize AI. The company uses Apache TVM to help companies optimize their neural nets for a wide variety of hardware.
Read More: zdnet