(65) Prior Publication Data US 2011/021928O A1 Sep. 8, 2011 Related US Application Data
(60) Division of application No. 1 1/572,372, filed as application No. PCT/US2005/025616 …

A system and method for enabling high-speed, low-latency global collective
communications among interconnected processing nodes. The global collective network …

(56) References Cited 2007/0226686 Al 9, 2007 Beardslee et al. 2007/0242611 A1 10,
2007 Archer et al. US PATENT DOCUMENTS 2007.0245122 A1 10, 2007 Archer et al …

US PATENT DOCUMENTS 4,715,032 A 12/1987 Nilsson 4,843,540 A 6, 1989 Stolfo 5,
101480 A 3, 1992 Shin et al. 5,105,424 A 4/1992 Flaig et al. 5,333,279 A 7/1994 Dunning …

Methods, apparatus, and computer program products are dis closed for executing a gather
operation on aparallel computer according to embodiments of the present invention. Embodi …

Methods, apparatus, and products are disclosed for perform ing an all-to-all exchange on n
number of data buffers using XOR swap operations. Each data buffer has n number of data …

(57) ABSTRACT A parallel computer that includes compute nodes having computer
processors and a CAU (Collectives Acceleration Unit) that couples processors to one …

Collective operation protocol selection in a parallel computer that includes compute nodes
may be carried out by calling a collective operation with operating parameters; selecting a …

US8565089B2 - Performing a scatterv operation on a hierarchical tree network optimized
for collective operations - Google Patents US8565089B2 - Performing a scatterv operation …

Methods, compute nodes, and computer program products are provided for DMA engine
assisted local reduction. Embodiments include receiving, by a DMA engine, one or more …