cudaCR: An In-Kernel Application-Level Checkpoint/Restart Scheme for CUDA-Enabled GPUs.
Behnam PourghassemiAparna ChandramowlishwaranPublished in: CLUSTER (2017)
Keyphrases
- application level
- operating system
- general purpose
- network management
- quality of service
- network services
- virtual machine
- overlay network
- gpu implementation
- random walk
- graphics hardware
- parallel programming
- parallel processing
- kernel methods
- graphics processors
- database
- bottle neck
- knowledge discovery
- feature space
- data analysis
- support vector
- web services