Author: Tristan Konolige and Sayce Falk
This post now lives at: https://octoml.ai/blog/in-the-cloud-sparsity-on-gpus-provides-5x-speedup/
Up to 9x performance improvements with TVM’s new auto-scheduler