As You see from this image, new SGI Octane III has 10 blades, each 2 x 4-core Intel 5500 series CPU’s, total 80 cores per on small rack. Pretty neat supercomputer. All Blades are interconnected by Gigabit Ethernet or InfiniBand. Supercomputer supports Red Hat or Suse, SGI Isle cluster manager and Altair PBS patch scheduler.
GT300 announcement is going to be in December – NVIDIA head during October will visit TSMC for checking the readiness degree of this video chip for mass deliveries according to several sources.
NVIDIA will simultaneously discuss the prices of 40 nm production , which is released by TSMC. During July the last company was possible to raise the release level of suitable 40 nm chips from 30% to 60%.
GT300 will be the most powerful CUDA chip ever. Based on technicad data about NVIDIA GT300 shader units frequencies, memory speed and other, people fom Hardware-Infos (Germany) have decided, that G300 performance will be 2,4 TFLOPS. It’s more than twice GTX280 cards 0,933 TFLOPS. Comparision is not direct, as G300 supports MIMD architecture, but GTX280 only SIMD. GPU is rumoured to have rather tremendous die area of 452mm², which means that only about 104 – 105 of such dies can fit onto one 300mm wafer.
Nvidia received first working samples from TSMC (Taiwan Semiconductor Manufacturing Company) in the beginning of September.
After long summer vacation I restarted my GPU cluster @ 166K PPD.
ca 165 000 - 166 000 PPD
William Dally, Nidia chief scientist
Yes, these are the figures said by Chief scientist of Nvidia Corporation., William Dally. Its 20 times more, than today best GPU’s lige GTX295 (about 1 TFLOPS) do. Key factors are MIMS technology, allowing to prosess multiple different instructions at same time. Also added asynchronous capabilities, making GPU’s very effective in parallel tasks.
Wiiliam Dally was presently the chairman of Stanford University’s computer science department.
Hmm. No meat, no fish, not even dessert this SuperMicro GPU 1U system. What is it really? Only 2 Tesla 1060 GPU cards and less than 2 TFLOPS of power?! Why all industrial/corporate companies hesitate to use double GPU cards? Cooling is the key, if You are able to keep GPU card temperature below 80 degrees C, It’s very stable. But double power!
If I’d have to make 1U massively parallel number-cruncher, it will be at least 6 TFLOPS thing. Three times over Supermicro How? By using 3 double-GPU GT200 cards side by side, more effective cooling by using radial fans (instead of funny small ones, used in SuperMicro), PCIe extender cables, small 1-socket motherboard. There is no need for multi-HDD RAID systems, so I’d leave HDD subsystem for 2,5″ or SSD.
Source: Harward University