3 ms·
Since you're using GH200s for these optimizations you're restricted to single device workloads (since GH series are SOC architecture). Kimi K2 (and many other l
by jakestevens2 7mo ago
Since you're using GH200s for these optimizations you're restricted to single device workloads (since GH series are SOC architecture). Kimi K2 (and many other large MoE models) requires multiple devices. Does that mean you can't scale these optimizations to multi-device workloads?
- 2uryaa 7mo agoHey Jack, we use GB200s for these workloads. Feel free to check those big models out on our site! We are doing Kimi, GLM, Minimax, etc.
- jakestevens2 7mo agoNice! But that doesn’t answer the question. Do these optimizations don’t scale to multi-device workloads or not?