Support for MPI at the Network Interface Level

Tourancheau, Bernard; Westrelin, Roland

doi:10.1007/3-540-45417-9_12

Cited by 5 publications

(3 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Unexpected messages are guaranteed to traverse the entire posted receive queue. This data explains results in [29], which indicate that the performance advantage of MPI offload decreases as the number of processors increases. In- The average search length for the unexpected queue is relatively short for all applications except FT. For FT, the search length grows linearly with the number of processors.…”

Section: Maximum Search Lengthsupporting

confidence: 57%

See 1 more Smart Citation

An analysis of NIC resource usage for offloading MPI

Brightwell

Underwood

18th International Parallel and Distributed Processing Symposium, 2004. Proceedings.

View full text Add to dashboard Cite

show abstract

Section: Maximum Search Lengthsupporting

confidence: 57%

“…For example, in [17] portions of a Portals [5] stack were offloaded. Similarly, the Quadrics network [20] offloads the MPI matching stack onto the network interface, and others have explored this approach for Myrinet [29]. A new software stack from Myricom for Myrinet, called MX [19], appears to do this as well.…”

Section: Related Workmentioning

confidence: 99%

An analysis of NIC resource usage for offloading MPI

Brightwell

Underwood

18th International Parallel and Distributed Processing Symposium, 2004. Proceedings.

View full text Add to dashboard Cite

show abstract

“…There is a mention of MPI message matching offloading on Quadrics [17] and Myrinet [23] network interfaces using on-NIC thread and memory. For InfiniBand [2], TupleQ [16] has been proposed where a Shared Receive Queue (SRQ) is created for each <contextId, rank, tag> tuple to match messages in hardware and improve the Rendezvous protocol [18].…”

Section: Introductionmentioning

confidence: 99%

An Efficient MPI Message Queue Mechanism for Large-scale Jobs

Zounmevo

Afsahi

2012

2012 IEEE 18th International Conference on Parallel and Distributed Systems

View full text Add to dashboard Cite

The Message Passing Interface (MPI) message queues have been shown to grow proportionately to the job size for many applications. With such a behaviour and knowing that message queues are used very frequently, ensuring fast queue operations at large scales is of paramount importance in the current and the upcoming exascale computing eras. Scalability, however, is two-fold. With the growing processor core density per node, and the expected smaller memory density per core at larger scales, a queue mechanism that is blind on memory requirements poses another scalability issue even if it solves the speed of operation problem. In this work we propose a multidimensional queue traversal mechanism whose operation time and memory overhead grow sub-linearly with the job size. We compare our proposal with a linked listbased approach which is not scalable in terms of speed of operation, and with an array-based method which is not scalable in terms of memory consumption. Our proposed multidimensional approach yields queue operation time speedups that translate to up to 4-fold execution time improvement over the linked list design for the applications studied in this work. It also shows a consistent lower memory footprint compared to the array-based design.

show abstract