对比：V4.1 Flash 每 GB200 每小时收入 38 美元，Kimi 3.1 Pro 为 8.34 美元
按每 GPU 每小时收入比较，DeepSeek V4.1 Flash 在 GB200 上为 38 美元，Kimi K3.1 Pro 在 MI355X 上为 8.34 美元，并称 Engram、因果编解码、KV 复用架构会被广泛采用。
The choice can not be clearer than this: 1) Kimi 3.1 Pro's 8.34$ income per AMD MI355X GPU / hr vs 2) DS V4.1 Flash' 38$ income per GB200 GPU / hr There is no match on economics between Moonshot and DeepSeek (although they are different model tiers). Off course, you can split hair and say it was on NVL72 for DS numbers. So what. Most GPUs are shipping in rack configurations / large scale up now. Nobody can avoid the pull of DS V4.1 Flash architecture: - Engram - Causal Encoder-Decoder (or Self Decoder) - KV Re-use In USA (SpaceX, Muse) and in China (Xiaomi, GLM, Kimi) all will have to adopt DS V4.1 Flash architecture. It is too good not to. Caveat: On Helios type rack/NVL72 Kimi K3.1's income per GPU will improve, though will be far less than that of DS v4.1 Flash.
这条是引用着另一条写的：x.com/semianalysis_/status/2114496296455184834