Session
MPI Optimization and Characterization
Event TypePaper
TP
Architectures
MPI
Networks
Performance
Programming Systems
State of the Practice
TimeWednesday, November 14th10:30am - 12pm
LocationC140/142
Presentations
| 10:30am - 11:00am | Cooperative Rendezvous Protocols for Improved Performance and Overlap | |
| 11:00am - 11:30am | Framework for Scalable Intra-Node Collective Operations Using Shared Memory | |
| 11:30am - 12:00pm | Characterization of MPI Usage on a Production Supercomputer |