Timezone: »

 
High Performance Hierarchical Tucker Tensor Learning Using GPU Tensor Cores
hao huang · Xiao-Yang Liu · Weiqin Tong · Tao Zhang · Anwar Walid

We present efficient HT tensor learning primitives using GPU tensor cores and demonstrate three applications,namely third-order and high-order HT decompositions, a HT tensor layer for deep neural network and a HT tensor-tree structure quantum machine learning algorithm TTN. Experimental results demonstrated that our optimized algorithms have higher efficiency.

Author Information

hao huang (Shanghai University)
Xiao-Yang Liu (Columbia University)
Weiqin Tong (Shanghai University)
Tao Zhang (Shanghai University)
Anwar Walid (Nokia Bell Labs)

More from the Same Authors