[SYSTEMDS-1034] Implement solve builtin function using cublas kernels

• Type: Sub-task
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Fix Version/s: SystemML 0.15
• Component/s: Runtime
• Assignee: Nakul Jindal
• Reporter: Niketan Pansare
• Created: 12/Oct/16 22:28
• Updated: 09/Sep/17 04:58
• Resolved: 24/May/17 05:08

Description

1. Extend BinaryOp to enable GPU for solve
2. Add MatrixMatrixBuiltinGPUInstruction and use JCuBlas2's cublasDtrsmBatched and cublasDgeqrfBatched (or cublasDgetrfBatched) methods.

For reference implementation, see https://github.com/apache/incubator-systemml/blob/master/src/main/java/org/apache/sysml/runtime/matrix/data/LibCommonsMath.java#L97

nakul02