TC-GNN: Bridging Sparse GNN Computation and Dense Tensor Cores on GPUs.
Yuke WangBoyuan FengZheng WangGuyue HuangYufei DingPublished in: USENIX Annual Technical Conference (2023)
Keyphrases
- high order
- tensor decomposition
- dense optical flow
- parallel processing
- general purpose
- parallel architectures
- high dimensional
- motion field estimation
- dimensionality reduction
- diffusion tensor
- tensor factorization
- dense stereo
- dense motion estimation
- real time
- parallel computers
- parallel programming
- linear algebra
- parallel implementation
- sparse representation
- higher order
- optical flow
- artificial intelligence