CUDA-DTM: Distributed Transactional Memory for GPU Clusters.
Samuel IrvingSui ChenLu PengCostas BuschMaurice HerlihyChristopher J. MichaelPublished in: NETYS (2019)
Keyphrases
- parallel computing
- commodity hardware
- parallel programming
- parallel implementation
- gpu accelerated
- graphics processors
- graphics hardware
- clustering algorithm
- gpu implementation
- transactional memory
- general purpose
- parallel execution
- parallel computation
- computing systems
- real time
- parallel architectures
- parallel processing
- massively parallel
- distributed environment
- distributed systems
- processing units
- data transfer
- hierarchical clustering
- cloud computing
- compute unified device architecture
- speculative execution
- message passing
- parallel algorithm
- fine grained