TOKENMAXXING: @vllm_project and @inferact achieve 1.3 BILLION tokens per $ of TCO (hyperscaler volume pricing) via disaggregated serving on AgentX. This result is 4x better than the B300 aggregated configuration.

更多来自 @SemiAnalysis_

1. #1

   Video posted by SemiAnalysis (@SemiAnalysis_)

   SemiAnalysisSemiAnalysis @SemiAnalysis_ · 19 小时前

   ▶
2. #2

   BREAKING NEWS🚨: @Inferact & Google submit their FIRST preliminary TPUv7 Kimi K3 2.8T Model Submission, and it is already delivering strong perf per TCO against GB200 NVL72, B3…

   SemiAnalysisSemiAnalysis @SemiAnalysis_ · 1 天前

   
3. #3

   BREAKING NEWS🚨: NVIDIA has just achieved 423 tok/s/user 🚀 on the 2.8T Parameter Kimi K3 model via a new upstream vLLM optimization. 1.68x Faster on the Same Hardware! This was don…

   SemiAnalysisSemiAnalysis @SemiAnalysis_ · 1 天前

   
4. #4

   GLM 5.3’s biggest leap came from post-training alone. “GLM 5.3 was built entirely by post-training the previous model. For me, that was a wake-up call.” “Post-training is far more…

   SemiAnalysisSemiAnalysis @SemiAnalysis_ · 1 天前

   ▶
5. #5

   MEME: If GPT-7 is AGI, OpenAI could potentially ditch Broadcom entirely and bring everything in-house, as GPT-7 could be better than Broadcom at designing 224G SerDes, handling PD…

   SemiAnalysisSemiAnalysis @SemiAnalysis_ · 2 天前

   
6. #6

   The agent access war has started！ Meta's Muse recently hit #1 on the App Store. But soon Amazon blocked it. Jefferies then found a third of the US shopping and travel sites it tes…

   SemiAnalysisSemiAnalysis @SemiAnalysis_ · 2 天前

   

查看 @SemiAnalysis_ 的全部帖子 →