Name
Abstract | ||
|---|---|---|
Message passing has been adopted as the main programming paradigm for many-core processors with on-chip networks for inter-core communication. To this end, message-passing libraries such as MPI can be used, as they provide well-known interfaces to application developers. Since MPI implementations were originally developed for macroscopic computer networks, the different characteristics of on-chip networks may require rethinking existing solutions. With the example of All reduce, we identify points where collective operations benefit from routines optimized for on-chip networks. The identified issues are then applied to additional collectives including Broadcast, All gather and All to all. The effectiveness of the proposed optimizations is demonstrated on the Single-Chip Cloud Computer (SCC), a many-core research chip created by Intel Labs. Experiments show that collective operations subjected to the identified optimizations are accelerated by factors roughly between 2 to 3 compared to current state of the art implementations. In addition to synthetic benchmarks, we show that the use of the optimized routines accelerates a scientific application by more than 40%. |
Year | DOI | Venue |
|---|---|---|
2012 | 10.1109/CLUSTER.2012.58 | Cluster Computing |
Keywords | Field | DocType |
intel scc,application developer,additional collective,proposed optimizations,optimized routine,scientific application,many-core processor,collective operation,many-core research chip,on-chip network,low-latency collectives,mpi implementation,synchronization,cloud computing,message passing,mpi,optimization,vectors,computer architecture,system on a chip,network on chip | Broadcasting,Computer architecture,System on a chip,Programming paradigm,Computer science,Parallel computing,Network on a chip,Implementation,Latency (engineering),Operating system,Message passing,Cloud computing | Conference |
ISSN | ISBN | Citations |
1552-5244 | 978-1-4673-2422-9 | 1 |
PageRank | References | Authors |
0.36 | 6 | 4 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Adan Kohler | 1 | 33 | 3.60 |
| Martin Radetzki | 2 | 223 | 26.33 |
| Philipp Gschwandtner | 3 | 58 | 7.15 |
| Thomas Fahringer | 4 | 2847 | 254.09 |