DeepSeek-R1 at 3,872 tokens / second on a single Nvidia HGX H200 (blogs.nvidia.com) 13 points by moondistance 1y ago ↗ HN
[–] billconan 1y ago ↗ https://news.ycombinator.com/item?id=42879864this is cerebras' 70B number, 1600 tokens / sec, not sure about the costs.
2 comments
[ 3.2 ms ] story [ 19.2 ms ] threadthis is cerebras' 70B number, 1600 tokens / sec, not sure about the costs.