First-Ever Measured Vera Rubin NVL72 Silicon Performance Stats
NVIDIA’s new Vera Rubin NVL72 racks represent a major leap in AI inference efficiency, and CoreWeave is one of the first cloud providers deploying them at scale. The key claim is that Vera Rubin can deliver up to 10× more tokens per megawatt compared to Blackwell, meaning dramatically higher output without increasing power consumption or datacenter footprint. This isn’t just a hardware refresh — it’s a fundamental shift in how efficiently large language models can run, especially for companies that rely heavily on inference workloads rather than training. CoreWeave’s early adoption positions them as a specialized NVIDIA‑native cloud optimized for high‑density GPU clusters and large‑scale LLM deployment. If the 10× efficiency improvement holds up in real‑world workloads, it could significantly lower inference costs, improve throughput, and make advanced AI models more accessible to both enterprises and startups. This announcement signals a potential reshaping of AI cloud economics, with
重要性：RSS 导入条目，等待编辑或代理补充市场影响解读。
My Bear Case for nVidia
Apologies for the stream of consciousness. It's almost 3AM and I've been benchmarking LLMs for the past 8 hours: Back when DeepSeek R1 came out, we saw the market react to the realization that a viable LLM doesn't need to be trained on the volume of cutting edge nVidia hardware you can only get on the white market. R1 was good. It wasn't the best, but it was a wake-up call for the west... that rhyme was not intentional... Over the past few days, we saw 2 insane releases from Chinese labs. Kimi, which was trained on a mix of nVidia and Chinese chips, crept up to Opus 4.8 territory. And Qwen3.8, which came out Monday morning (Chinese time) and performs similarly to Kimi, and it was trained exclusively on domestic Chinese hardware -- proving nVidia's not even needed to train a SotA model. Even domestically (in the US), Everyone and their mother has been working on training and inference ASICs and infra which will eventually displace nVidia hardware. Jensen's not dumb, so he's trying to ke
重要性：RSS 导入条目，等待编辑或代理补充市场影响解读。