On benchmarking collective MPI operations

DOI10.1007/3-540-45825-5_43MaRDI QIDQ4412054zbMATH OpenFDO

Authors Thomas Worsch, Ralf H. Reussner, Werner Augustin

Publication date 13 July 2003

Published in Lecture Notes in Computer Science (Search for Journal in Brave)

Full work available at URL http://link.springer.de/link/service/series/0558/bibs/2474/24740271.htm

Mathematics Subject Classification ID

Performance evaluation, queueing, and scheduling in the context of computer systems (68M20) Computing methodologies and applications (68U99)

Recommendations

• Benchmarking collective operations with SKaMPI
• Accurately measuring overhead, communication time and progression of blocking and nonblocking collective operations at massive scale
• Towards an accurate model for collective communications
• Challenges and successes in achieving the potential of MPI
• An MPI benchmark program library and its application to the Earth Simulator

Cited in

(7)

• Performance evaluation and modeling of collective communications on Cray T3D
• Towards an accurate model for collective communications
• Accurately measuring overhead, communication time and progression of blocking and nonblocking collective operations at massive scale
• Potential performance improvement of collective operations in UPC
• Benchmarking collective operations with SKaMPI
• Understanding performance of SMP clusters running MPI programs
• Effective communication and file-I/O bandwidth benchmarks

This page was built for publication: On benchmarking collective MPI operations

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4412054)