Evaluating Performance Portability of OpenMP for SNAP on NVIDIA, Intel, and AMD GPUs Using the Roofline Methodology.
Neil A. MehtaRahulkumar GayatriYasaman GhadarChristopher KnightJack DeslippePublished in: WACCPD@SC (2020)
Keyphrases
- graphics processing units
- parallel programming
- multi core processors
- general purpose
- parallel processing
- gpu implementation
- parallel computing
- high performance computing
- graphics hardware
- computer architecture
- real time
- graphics processors
- shared memory
- massively parallel
- neural network
- computing systems
- floating point
- processing speed
- design methodology
- programming environment
- processing units
- parallel implementation
- parallel execution
- highly parallel
- efficient implementation
- times faster
- parallel algorithm
- heterogeneous computing
- general purpose computing