3 ms·
The computer as a whole has an entry (3.) in the top500 list. And then the cpu-only part of the computer has another entry (84.). The whole computer does about
by sampo 4y ago
The computer as a whole has an entry (3.) in the top500 list. And then the cpu-only part of the computer has another entry (84.). The whole computer does about 150 PFlop/s, and the cpu-only part about 6 PFlop/s. So 96% of the computing power comes from the GPU cards.
https://www.top500.org/lists/top500/list/2022/06/ https://www.top500.org/lists/top500/list/2022/06/
- credit_guy 4y agoThis is impressive, but not exceedingly so. A single NVidia H100 GPU gets 1 PFlops/s [1] for single float (32 bit) precision. There's an asterisk which says Shown with sparsity. Specifications are one-half lower without sparsity I have to admit that I don't understand what this means, but the apparent takeaway is that the "non-sparse" speed is 0.5 PFlops/s. So this supercomputer has a speed equal to about 300 H100 GPUs. Not bad, but not something where one would start worrying about renewable electricity or not. [1] https://www.nvidia.com/en-us/data-center/h100/ https://www.nvidia.com/en-us/data-center/h100/
- ajtulloch 4y agoTF32 is not IEEE-754 float32, it is a reduced precision format designed for machine learning usecases. The correct specsheet number for FP32 (and FP64 which is the relevant precision here) throughput on H100 is more like 60TFLOP/s, so your number is off by roughly an order of magnitude.