Optimizing MPI Communication on Multi-GPU Systems Using CUDA Inter-Process Communication.
Sreeram PotluriHao WangDevendar BureddyAshish Kumar SinghCarlos RosalesDhabaleswar K. PandaPublished in: IPDPS Workshops (2012)
Keyphrases
- communication protocol
- communication networks
- computer systems
- neural network
- computer networks
- distributed systems
- complex systems
- real time
- computing devices
- general purpose
- parallel processing
- management system
- communication cost
- communication channels
- communication technologies
- graphics processing units
- multithreading
- single point of failure