Ken Lloyd wrote: > With a little reorganization and forethought, you can even have your own > mini-supercomputer using banks of GPU cards to crunch vectors and matrices. > See Nvidia's CUDA development system, and their Tesla computer system. > This seems like one area where Clearspeed would really shine. Given artificial neurons with a fixed number of inputs and outputs.
http://www.clearspeed.com/docs/resources/ClearSpeed_e710_Marketing_Brochure_5-08.pdf At 3.84 GFLOPS/watt it's nearly ten times more efficient than Roadrunner @ 473 MFLOPS/Watt. And Roadrunner being one of the most energy efficient supercomputers.. http://www.green500.org/lists/2008/06/ranks1-100.php Marcus ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org
