Heterogeneous programming using OpenMP and CUDA/HIP for hybrid CPU-GPU scientific applications.
Marc GonzálezEnric MoranchoPublished in: Int. J. High Perform. Comput. Appl. (2023)
Keyphrases
- graphics processing units
- compute unified device architecture
- gpu implementation
- general purpose
- graphics processors
- parallel implementation
- parallel programming
- parallel computing
- graphics hardware
- parallel computation
- cpu implementation
- shared memory
- parallel processing
- real time
- computing systems
- high performance computing
- massively parallel
- efficient implementation
- floating point
- general purpose computing
- processing units
- parallel execution
- parallel architectures
- programming environment
- message passing
- scientific computing
- programming language
- commodity hardware
- data intensive