MICRO'23 TorchSparse++: Efficient Training and Inference Framework for Sparse Convolution on GPUs 

Talk video for MICRO 2023 paper: "TorchSparse++: Efficient Training and Inference Framework for Sparse Convolution on GPUs" (Oct 30th at Westin Harbour Castle, Toronto, ON, Canada)
Haotian Tang*, Shang Yang*, Zhijian Liu, Ke Hong, Zhongming Yu, Xiuyu Li, Guohao Dai, Yu Wang, Song Han
For more info, please visit:
TorchSparse website: torchsparse.mi...
Project website and paper: hanlab.mit.edu...
TorchSparse open-source: github.com/mit...