Performance portable GPU code generation for matrix multiplication.
Toomas RemmelgThibaut LutzMichel SteuwerChristophe DubachPublished in: GPGPU@PPoPP (2016)
Keyphrases
- code generation
- matrix multiplication
- application development
- software development
- code generator
- modeling language
- lightweight
- model driven
- software reuse
- formal specification
- rapid prototyping
- message passing
- parallel implementation
- design patterns
- parallel computing
- parallel computation
- distributed memory
- case study
- parallel algorithm
- software engineering
- dynamic programming