{"data":{"slug":"nvidia-vera-rubin-nvl72","etype":"server","name":"NVIDIA Vera Rubin NVL72","cat":"Rack-Scale System","subtitle":"The third-generation NVLink rack: 72 Rubin GPUs and 36 Vera CPUs in the Oberon frame.","summary":"Vera Rubin NVL72 is NVIDIA's 2026 rack-scale system and the successor to GB300 NVL72. It keeps the Oberon rack layout of 72 GPU sockets and 36 CPU sockets, filled with Rubin GPUs, Vera CPUs, NVLink 6 switches, ConnectX-9 SuperNICs and BlueField-4 DPUs. NVIDIA's spec table rates the rack at 3,600 PFLOPS of NVFP4 inference (sparse) with 20.7 TB of HBM4. First styled NVL144, counting GPU dies, it was restyled NVL72 to count packages. Production shipments began in August 2026; CoreWeave and Oracle Cloud made it available in September and October 2026.","attrs":[{"b":"Vendor spec","g":"Overview","l":"Configuration","v":"72 Rubin GPUs + 36 Vera CPUs (36 Vera Rubin superchips)"},{"b":"Vendor spec","g":"Memory","l":"HBM4","n":"NVIDIA spec table (retrieved 16 Nov 2026)","v":"20.7 TB per rack; 1,400 TB/s"},{"b":"Vendor spec","g":"Memory","l":"CPU memory","v":"up to 54 TB LPDDR5X"},{"b":"Vendor spec","g":"Compute","l":"NVFP4 inference","n":"Sparse, rack","u":"PFLOPS","v":"3,600"},{"b":"Vendor spec","g":"Compute","l":"NVFP4 training","n":"Dense, rack","u":"PFLOPS","v":"2,520"},{"b":"Vendor spec","g":"Compute","l":"FP8/FP6 training","n":"Dense, rack","u":"PFLOPS","v":"1,260"},{"b":"Vendor spec","g":"Overview","l":"NVLink 6 domain","n":"NVIDIA spec table (retrieved 16 Nov 2026); the January 2026 launch release gave 3.6 TB/s per GPU and 260 TB/s per rack","v":"All 72 GPUs at 3 TB/s each: 216 TB/s aggregate"},{"b":"Vendor spec","g":"Overview","l":"CPU cores","v":"3,168 Olympus cores, 6,336 threads per rack"},{"g":"Power","l":"Rack power","v":"Not publicly disclosed"},{"b":"Reported","g":"Overview","l":"Naming","n":"Next Platform, Mar 2026","v":"Styled NVL144 at GTC 2025 (die count); NVL72 from 2026 (package count)"},{"b":"Official","g":"Overview","l":"Availability","v":"Production shipments from Aug 2026; available on CoreWeave 30 Sep 2026; GA on OCI 27 Oct 2026"}],"kf":{"chips":"72 × Rubin + 36 × Vera","maker":"NVIDIA + ODMs","power":"Not publicly disclosed"},"aliases":["Vera Rubin NVL72","VR200 NVL72","Vera Rubin NVL144"],"ticker":null,"exchange":null,"hq":null,"founded":null,"website":null,"status":"Shipping","region":null,"asof":"2026-11-16","sources":[{"url":"https://www.nvidia.com/en-us/data-center/vera-rubin-nvl72/","label":"NVIDIA Vera Rubin NVL72 specifications"},{"url":"https://investor.nvidia.com/news/press-release-details/2026/NVIDIA-Kicks-Off-the-Next-Generation-of-AI-With-Rubin--Six-New-Chips-One-Incredible-AI-Supercomputer/default.aspx","label":"NVIDIA Rubin platform launch (CES, Jan 2026)"},{"url":"https://investor.nvidia.com/news/press-release-details/2026/NVIDIA-and-Oracle-Expand-NVIDIA-Vera-Rubin-to-Multi-Gigawatt-Scale-and-Bring-NVIDIA-Vera-CPU-to-Oracle-Cloud-Infrastructure/default.aspx","label":"NVIDIA and Oracle: Vera Rubin GA on OCI (Oct 2026)"},{"url":"https://www.coreweave.com/news/coreweave-delivers-nvidia-vera-rubin-nvl72-performance-at-production-scale-starting-with-cognition","label":"CoreWeave: Vera Rubin NVL72 in production (Sep 2026)"},{"url":"https://www.nextplatform.com/compute/2026/03/19/driving-down-the-ai-system-roadmap-with-nvidia/5210195","label":"Next Platform: GTC 2026 roadmap (Mar 2026)"}]},"meta":{"found":true,"count":1,"served_from":"site cache (no live upstream calls)","as_of":"2026-11-16T22:03:33+00:00","disclaimer":"Research and information only, not investment advice. Structured records carry basis labels (vendor spec / reported / estimate); market quotes are delayed/indicative.","attribution":"Records compiled from vendor documentation, official filings and reputable public reporting; market quotes via financial data providers. See /data-sources."}}