SemiAnalysis：英偉達VR NVL72與GB200 NVL72推理TCO成本對比與架構分析

《SemiAnalysis：海思麒麟9030及其N+3制造工藝深度逆向工程分析報告》

《SemiAnalysis最新報告：英偉達VR NVL72與GB200 NVL72推理TCO成本對比與架構分析》

《2025~2026 OCP/FMS/ISSCC/ODCC/HotChips全球峰會合集》

《2026年CPU處理器技術合集》 

《2026/2025超節點數據中心峰會合集》

《335+份DeepSeek技術報告合集》

《AI智算細分技術深度分析合集》

《AI技術通識&基礎合集》

《2026年華為韜（τ）定律合集》
《800+份重磅ChatGPT/AIGC專業報告》，《智算中心發展趨勢報告（2026）》已傳至智能計算芯知識，請批量下載。

Vera Rubin NVL72 是英偉達機架級Oberon 架構的第二代產品，其在推理性能上的提升源于極致的協同設計。工程樣品的初步結果令人鼓舞。搭載 DeepSeek R1 的Vera Rubin NVL72，相比目前的 GB200 NVL72，每兆瓦性能提升 5.4 倍，每美元性能提升 5 倍；而在 2025 年 GB200 NVL72 早期調試階段，兩者的性能差距甚至更大。目前Vera Rubin 仍處于早期調試階段，我們預計這一差距還將繼續擴大。隨著軟件生態的成熟，Rubin 的推理性能將持續提升，這與我們在InferenceX 基準測試中為Blackwell 展示的規律一致，Rubin 仍有廣闊的發展空間。

Nvidia has also recently made available their first public release of the Rubin (SM_107) software stack with CUDA13.4 and has upstreamed Rubin PRs to PyTorch, vLLM and OpenAI Triton Compiler. Blackwell was not able to reuse Hopper WGMMA kernels, Rubin is able to reuse Blackwell's kernels, which makes the software bring up process much smoother. For speed of light (SOL) performance, engineers will still need to tune and rewrite kernels but for those that are focused on time to market, Blackwell kernels can be reused. We will also explain Rubin’s new 3 bit programmable LUT tensor core. 英偉達近期也推出了搭載 CUDA13.4 的 Rubin（SM_107）軟件棧的首個公開版本，并將 Rubin 的拉取請求同步到了 PyTorch、vLLM 以及 OpenAI Triton 編譯器中。Blackwell 架構無法復用 Hopper 架構的 WGMMA 內核，而 Rubin 架構能夠復用 Blackwell 的內核，這使得軟件適配過程順暢得多。在光速性能方面，工程師仍需對內核進行調優和重寫，但對于那些注重上市時間的開發者而言，Blackwell 內核可以直接復用。我們還將介紹 Rubin 全新的 3 位可編程查找表張量核心。

NVIDIA has also released on GitHub that Feynman is SM_140. Unlike Blackwell to Rubin, Rubin to Feynman will be an much more complex transition on the kernel front.NVIDIA 也在 GitHub 上發布了 Feynman 對應的架構版本為 SM_140。與 Blackwell 到 Rubin 的過渡不同，Rubin 到 Feynman 在核心層面將是一個復雜得多的過渡。

The early metrics gathered on VR NVL72 come from CoreWeave. We have not independently verified them. Nvidia has committed to submitting verifiable numbers to InferenceX by Q3 CY2026. Google should submit TPUv7 results in the next couple of months, and AMD has committed to MI455X UALoE72. Once those land, the ecosystem gets an objective comparison across systems.關于VR NVL72的早期性能指標來自CoreWeave公司，我們尚未對其進行獨立驗證。英偉達已承諾在2026年第三季度前向InferenceX提交可驗證的數據。谷歌應在未來幾個月內提交TPUv7的相關結果，而AMD也承諾推出MI455X UALoE72產品。一旦這些數據和產品落地，整個生態系統將能實現不同系統間的客觀對比。

In this article we break down Nvidia's Rubin claims against several baselines, showing where Rubin clearly leads Blackwell and where the lead is thinner. We will also analyze Rubin’s performance per total cost of ownership using our already existing estimates for Rubin’s total cost of ownership (TCO). The TCO for Rubin and many other systems is sourced from our AI TCO model, which tracks the total cost of ownership of different AI chips, factoring in capex, opex and different other expenses. We also consider performance per watt using our All-in Utility Provisioned Power Estimates from our Datacenter Model.在本文中，我們詳細分析了英偉達Rubin相較于多個基準的性能表現，明確指出Rubin在哪些方面顯著領先于Blackwell，以及在哪些方面的優勢相對有限。我們還將結合此前對Rubin總擁有成本（TCO）的估算，分析Rubin的單位總擁有成本性能。Rubin及其他諸多系統的總擁有成本數據均源自我們的AI總擁有成本模型，該模型會考量資本支出、運營支出及其他各類費用，對不同AI芯片的總擁有成本進行追蹤。此外，我們還將依據數據中心模型中的全功能配置功耗估算值，分析Rubin的單位功耗性能。

Finally, we will present a component by component build up of the Bill of Materials (BoM) for the VR NVL72. This is available in our upcoming SemiAnalysis Bill of Materials (BoM) Model.最后，我們將為 VR NVL72 逐步拆解展示物料清單（BoM）的各個組成部分。相關內容可在我們即將推出的 SemiAnalysis 物料清單（BoM）模型中查看。

level.Another area where Rubin Oberon NVL72 will fare better than Blackwell Oberon NVL72 is in a much faster production ramp period. This is thanks to Rubin’s simpler cableless compute tray design and learnings from Nvidia’s experience with deploying a rack-scale copper backplane, having invested much effort into ironing out issues with Blackwell’s copper backplane. Our Accelerator Model tracks quarter by quarter shipments of Rubin at both the package level and the rack level.Rubin Oberon NVL72 相比 Blackwell Oberon NVL72 表現更出色的另一個領域是生產爬坡周期快得多。這得益于 Rubin 更簡潔的無纜計算托盤設計，以及英偉達在部署機架級銅背板方面的經驗積累——英偉達已投入大量精力解決了 Blackwell 銅背板的相關問題。我們的加速器模型按季度追蹤 Rubin 封裝級和機架級的出貨量。

本報告內容來自“積硅步-集大成”，以上僅為一小部分內容，僅限于學習研究用途，Word/PDF下載鏈接“《SemiAnalysis報告：英偉達VR NVL72與GB200 NVL72推理TCO成本對比與架構分析》”。

本號知識合集

免責申明：本號聚焦相關技術分享，內容觀點不代表本號立場，可追溯內容均注明來源，發布文章若存在版權等問題，請留言聯系刪除，謝謝。