| Document Date: 2010-05-06 00:14:41
Open Document File Size: 333,81 KBShare Result on Facebook
City Nice / / Company IBM / Mellanox / Optimizing Bandwidth Limited / / Country France / / / Event Natural Disaster / / Facility University of Tennessee / G. Pike / University of California / / IndustryTerm higher overall efficiency on systems / interconnect hardware / unordered network hardware / cluster networking hardware / shared memory hardware / scientific applications / cross-processor / interconnect technologies / bandwidth-sensitive applications / eager algorithm / host processor / distributed algorithm / message passing systems / communication protocol / parallel vector systems / collective communications / compiler run-time systems / lower software / parallel systems / exhaustive search / production applications / cluster networks / communication startup / target networks / cluster network hardware / given network / Message strip mining heuristics / target processor / lower software overhead / communications libraries / speed networks / cluster hardware / realworld scientific applications / recent interconnect technologies / performance networks / processor-based solutions / computing / parallel algorithm / communication network / remote processors / near-commodity network / rendezvous algorithm / / MarketIndex NAS FT / / OperatingSystem UNIX-like / / Organization University of California / Berkeley / Berkeley National Lab / University of Tennessee / Knoxville / / Person J. Duell / Elan Figure / D. Bonachea / K. Datta / C. Iancu / K. Yelick / / Position representative for one-sided communication / Governor / high performance programming model for clusters / representative / programmer / / Product Slabs / GASNet / One-Sided Communication / / ProgrammingLanguage C / Fortran / / ProvinceOrState Tennessee / California / / Technology Alpha / LAN / 3-D / UNIX / API / NIC processor / parallel algorithm / eager algorithm / ATM / 64 processors / UDP / RDMA-based interconnect technologies / shared memory / operating systems / recent interconnect technologies / increasingly offloading communication protocol / Slabs algorithms / Ethernet / target processor / rendezvous algorithm / host processor / Pencils algorithm / MPI protocol / Exchange algorithms / 512 processors / Parallel Processing / / URL http /
SocialTag |
mflops-upc-slabs_vs_pencils.eps