[SYSTEMDS-447] Phase 2: Custom GPU kernels for complex operator patterns and for improving performance of sparse operators - ASF JIRA

XML

Word

Printable

JSON

1.	Improve the performance of copyUpperToLowerTriangleDense kernel	Open	Unassigned
2.	Chose the block dimensions of custom kernels using CUDA's occupancy API	Open	Unassigned
3.	Add custom kernels for dense binary elementwise operations	Open	Unassigned
4.	Add custom kernels for sparse binary elementwise operations	Open	Unassigned
5.	Add custom kernel for dense ternary aggregate instruction	Open	Unassigned
6.	Add custom kernel for sparse ternary aggregate instruction	Open	Unassigned
7.	Add support for aggregate unary operations on GPU	Open	Nakul Jindal
8.	Implement custom kernel for append operation	Open	Unassigned
9.	Implement custom kernel for reshape operation	Open	Unassigned
10.	Add bias_multiply operator	Closed	Niketan Pansare
11.	Add custom kernel for sparse matrix dense vector multiplication	Open	Unassigned
12.	Add custom kernel for sparse matrix dense matrix multiplication	Open	Unassigned