TokenPost topic
Vera Rubin
29 TokenPost stories are tagged Vera Rubin. The newest appear first.
Latest Vera Rubin stories
- Nvidia’s Vera Rubin Platform Generates Nearly $20 Billion in First Quarter of Mass-Production Shipments
Nvidia posted $115.8 billion in FY2027 Q3 revenue, up 103% year over year. Data center revenue reached $107.6 billion.
· Business
- Nvidia Ships Vera Rubin NVL72 With 72 GPUs and 36 CPUs
The platform delivered up to twice its predecessors’ per-GPU performance as of Nov. 16. Earlier this fall, AI executives raised concerns about cyberattacks and model control.
· Technology
- Nvidia outlook beats expectations as longer-term AI demand remains in focus
Max Weinbach said Nvidia’s next-quarter outlook exceeded his expectations, while longer-term computing demand and supply chain relationships remain uncertain.
· Insights & Views
- NVIDIA Expands Commitments With Six Partners for AI and Vera Rubin
The work will broaden manufacturing capacity for substrates, advanced packaging and wafers.
· Business
- Nvidia Expects Rubin HPC GPU to Become Available in 2028
The single-GPU processor will have more than 50,000 CUDA cores, while Nvidia expands its CUDA-X tools for scientific research.
· Technology
- Nvidia’s Vera Rubin Posts Fastest Times in Six MLPerf Training Tests
Its 1,024-GPU DeepSeek-V3 run finished in 4.14 minutes, with about twice GB300’s per-GPU performance at the same scale.
· Technology
- Industry Sources Raise HBM4 Concerns Over NVIDIA Vera Rubin Ramp
Suppliers of 800-volt power systems, M9 server-board materials and optical components could face changing demand expectations if the production ramp shifts.
· Technology
- South Korea Expects Nvidia GPU Deliveries Under Contracts to Continue
About 13,100 Nvidia B200 GPUs have been deployed in the country, and plans to introduce Vera Rubin GPUs are expected to proceed as planned.
· Technology
- Outer Ninth Circulates Deleted Message on Vera Rubin Supply Chain
The anonymous message did not identify the supplier; 生益科技 appeared only in market speculation, while 台光電’s material was cited as a comparison.
· Technology
- Nvidia Says Vera Rubin Reaches Up to 30 Times GB300’s Cost-Adjusted Inference Throughput
Vera Rubin NVL72’s per-GPU high-bandwidth memory (HBM) capacity is about 3% higher than GB300 NVL72’s. GB300 NVL72 reached up to 10 times H200’s cost-adjusted inference throughput on specific workloads.
· Technology
- Nvidia Says It Ships More Than 1,000 MGX Racks a Week
The modular systems are designed for power densities above 120 kilowatts per rack and can accommodate future Vera Rubin platforms.
· Technology
- Foxconn Subsidiary Becomes NVIDIA DSX Pod Global Partner
Foxconn’s initial share of NVIDIA’s Vera Rubin orders is expected at 40% to 45%. The potential NVL72 rack market is estimated at 96,000 units in 2027.
· Business
- Oracle and NVIDIA Expand Vera Rubin Deployment to Multi-Gigawatt Scale
Oracle Cloud Infrastructure will add NVIDIA Vera CPUs. Oracle separately partnered with Cerebras to use its equipment in Oracle AI data centers.
· Technology
- Vera Rubin Inference Results Leave Nvidia’s 75% Training Claim Untested
At about 190 output tokens per second per user, Vera Rubin reached 5.7 times GB300’s total throughput per megawatt of GPU and memory power, using a four-GPU GB300 comparison.
· Technology
- HUMAIN Plans Up to 84 Megawatts of AI Computing Capacity
The infrastructure is roughly equivalent to 27,000 Nvidia GPUs. Blackwell Ultra chip deployment is planned to begin in the fourth quarter of 2026, followed by Vera Rubin-series chips in 2027.
· Technology
- Supermicro Unveils 256-CPU Liquid-Cooled Rack for Agentic AI
The system is designed to work alongside NVIDIA’s Vera Rubin platform and support large-scale agentic AI applications.
· Technology
- Nvidia Modular AI Design Targets Deployment in as Little as 14 Days
The design moves system integration and validation to partner factories before delivery. Worldwide deliveries are set to begin in early 2027.
· Technology
- WPP Tests Google Cloud A5X on Workloads That Usually Need a Cluster
Google Cloud plans general availability for its A5X bare-metal GPU instances in November 2026.
· Technology
- NVIDIA Moves Vera Rubin AI Factory Testing Into Modular Pods
The design shifts system assembly and checks to manufacturing. Worldwide deliveries are planned for early 2027.
· Technology
- Google Cloud Plans Vera Rubin Clusters With Up to 960,000 GPUs
A single site could host up to 80,000 GPUs, while the 960,000 figure covers a connected cluster spanning multiple locations. General availability is planned for November 2026.
· Technology
- Google Cloud and NVIDIA Make Vera Rubin Infrastructure Broadly Available
The companies announced the availability on Oct. 21.
· Technology
- Nvidia Says DSX POD Could Cut AI Factory Deployment to 14 Days
The Vera Rubin reference design moves integration and testing into modular containers, with partners planning worldwide deliveries in early 2027.
· Technology
- NVIDIA Brings Modular AI-Factory Systems to OCP Summit
Vera Rubin combines five purpose-built racks, while MGX supports hundreds of configurations for AI computing, networking, storage, power and cooling.
· Technology
- AI Data Center Construction Costs Reach Up to $39.5 Billion per Gigawatt
Estimated capital spending ranges from $34.6 billion to $39.5 billion, with Nvidia’s Vera Rubin systems at the high end and OpenAI’s Jalapeno ASIC architecture lower.
· Technology
- NVIDIA, Eli Lilly Plan Up to $1 Billion for AI Drug Lab
The five-year partnership will link Lilly’s wet labs with computational systems and use NVIDIA’s BioNeMo platform and planned Vera Rubin architecture.
· Business
- SpaceXAI Plans Nvidia AI Expansion With Starmind Satellite
The plan includes Nvidia’s Vera Rubin platform and an orbital Starmind AI system, while SpaceX separately completed a $25 billion senior-notes offering in June.
· Technology
- CoreWeave Puts NVIDIA Vera Rubin NVL72 Into Production for Cloud Customers
Cognition’s early tests found up to 4.8 times the total token throughput of NVIDIA’s GB200 NVL72 in software-engineering AI inference tasks.
· Technology
- Supermicro, SpaceXAI Plans Point to Gigawatt-Scale AI Buildout
Supermicro’s Vera Rubin systems are designed to scale from 5 megawatts to gigawatt capacity as SpaceXAI expands infrastructure for Grok.
· Technology
- Supermicro Introduces Blueprint for Eight-Rack NVIDIA Vera Rubin System
The proposed design includes up to 1,152 Rubin GPUs, 576 Vera CPUs and 3.2 megawatts of power, with availability expected in the fourth quarter of 2026.
· Technology