Top
Author
Published
12 Oct 2026Form Number
LP2549PDF size
6 pages, 435 KBSubscribed to LP2549.
Thank you for your feedback.
Table of Contents
Introduction
At OCP Global Summit 2026, Lenovo is showcasing the latest expansion of Lenovo AI Cloud Gigafactory with the NVIDIA Vera Rubin Platform. Built on the complete NVIDIA Vera Rubin platform, the integrated, rack-scale solution is designed to help cloud service providers (CSPs) and neoclouds rapidly deploy production-ready infrastructure for advanced AI training and inference, turning growing AI demand into revenue-generating capacity.
Lenovo AI Cloud Gigafactory with NVIDIA
For AI cloud providers, the challenge is no longer simply securing high-density GPU infrastructure. Success depends on orchestrating increasingly complex infrastructure requirements to minimize execution risk, accelerate deployment and reach revenue-generating capacity faster.
The Lenovo AI Cloud Gigafactory with NVIDIA addresses that challenge through a comprehensive services-led model. The offering bridges Lenovo engineering expertise, global manufacturing, in-house integration and validation, Lenovo Neptune® Liquid Cooling experience and lifecycle services with NVIDIA-accelerated infrastructure, so providers can move from strategy to revenue-ready AI capacity with greater confidence.
"The AI race is no longer just about deploying GPUs. It's about how quickly providers can transform power, space and infrastructure investments into operational AI capacity and revenue," said Conor Malone, Vice President & GM, Cloud Service Providers, Lenovo. "Lenovo AI Cloud Gigafactory with NVIDIA brings together advanced AI infrastructure with Lenovo’s global manufacturing and deployment expertise, including capabilities at our North Carolina Smart Campus, to help customers build AI factories faster, scale with confidence, and unlock business value sooner."
“The next generation of AI cloud will require full-stack, rack-scale infrastructure designed for increasingly complex reasoning, inference and agentic AI workloads,” said Chris Marriott, Vice President of Enterprise Platforms and Solutions at NVIDIA. “By bringing the NVIDIA Vera Rubin Platform to Lenovo AI Cloud Gigafactory, Lenovo is helping cloud providers reduce deployment complexity, scale with confidence and turn infrastructure investments into revenue-ready AI capacity faster.”
Five Rack-Scale Systems, Delivered as One
As AI demand shifts from training to high-volume inference, reasoning and agentic AI, the unit of AI compute is moving from the individual rack to the full pod. Lenovo's NVIDIA Vera Rubin Platform brings together all five rack-scale systems of the NVIDIA Vera Rubin platform, aligned with NVIDIA DSX AI Factory Platform reference architecture guidance:
- NVIDIA Vera Rubin NVL72: Combines 72 Rubin GPUs and 36 NVIDIA Vera CPUs with NVIDIA NVLink 6 scale-up networking, NVIDIA ConnectX-9 SuperNICs and NVIDIA BlueField-4 DPUs. NVIDIA reports this system can achieve up to 30x higher throughput per megawatt, which directly lowers inference costs by up to 45x per million tokens - all while utilizing one fourth the GPUs to deliver AI intelligence when compared against previous generations.
- NVIDIA Groq 3 LPX: A dedicated inference accelerator rack for the low-latency, large-context demands of agentic systems, with 256 LPUs, 128GB of on-chip SRAM, 640 TB/s of scale-up bandwidth, NVIDIA ConnectX-9 based fabric expansion logic for scale-out networking, and NVIDIA BlueField-4 DPUs for scale-in networking.
- NVIDIA Vera CPU Rack: A dense, liquid-cooled NVIDIA MGX-based system with 256 Vera CPUs and 128 BlueField-4 DPUs. NVIDIA MGX is a modular reference architecture for accelerated computing, helping system builders design future-compatible systems faster, reducing costs, accelerating time to market, and improving ROI. A single rack can run more than 22,500 concurrent reinforcement learning or agent sandbox environments.
- NVIDIA CMX Context Memory Storage Platform: AI-native, rack scale context memory storage built on NVIDIA STX and powered by BlueField-4 STX storage processors. CMX extends GPU memory with a shared, pod-level tier optimized for KV cache.
- NVIDIA Spectrum-6 SPX Ethernet rack: High-performance networking racks for connecting compute within the Vera Rubin POD.
Manufacturing at Global Scale, with U.S. Production in North Carolina
Gigafactory-scale AI depends on the ability to build, integrate and validate liquid-cooled racks at volume. Lenovo's global/local manufacturing strategy spans more than 30 factories in 11 markets, pairing global scale with regional production close to customers. Lenovo's supply chain was also recently ranked fifth globally in the Gartner® Supply Chain Top 25 for 2026, its highest-ever placement.
In North America, that scale is anchored by the Lenovo North Carolina Smart Campus (NCSC) in Whitsett, North Carolina. Following a recent expansion, the broader campus spans 890,000 square feet, with NCSC encompassing 520,000 square feet of that footprint and infrastructure ready for 20 megawatts of power. It is future-proofed for the next two generations of high-performance computing, and Lenovo has added 35% more manufacturing capacity, including Direct Liquid Cooling. Located less than an hour from Lenovo's development labs, NCSC brings production closer to customers for faster delivery, greater flexibility and improved responsiveness.
Figure 3. Lenovo North Carolina Smart Campus (NCSC) in Whitsett, N.C. in the USA
Availability
Lenovo is now shipping NVIDIA Vera Rubin NVL72 globally, enabling CSPs, neoclouds, and AI infrastructure providers to accelerate AI factory deployments today, with early access to the broader NVIDIA Vera Rubin Platform as the ecosystem becomes available.
To further simplify deployment and growth, Lenovo Global Financial Services (LGFS) offers flexible financing and consumption models that help customers preserve capital, accelerate capacity expansion, and align infrastructure investments with business growth.
Lenovo AI Cloud Gigafactory at 2026 OCP Global Summit
Lenovo is showcasing the NVIDIA Vera Rubin Platform at Booth D35 during the 2026 OCP Global Summit, held October 12–15 at the San Jose Convention Center. Lenovo experts are on site to discuss AI Cloud Gigafactory design, liquid cooling, manufacturing and deployment for AI cloud builders.
Lenovo is also hosting two key speaking sessions:
- From Blueprint to Revenue: Engineering and Executing AI Cloud Gigafactories with NVIDIA
Monday. October 12, 1:15pm PT at SJCC Concourse Level 211. Featuring John Donovan, ED & GM of AI & Scale Cloud, Lenovo, and Chris Marriott, VP of Enterprise Platforms and Solutions, NVIDIA, will explore how the principles behind gigawatt-scale AI cloud buildouts translate into faster time to revenue for providers and enterprises alike.
- CLARA: Automated AI Cluster Architecture Design Tool
Tuesday, October 13, 1:45pm PT at SJCC Concourse Level 220B. Lenovo's Rainy Yan, AI Cluster Automation Engineer, and Reece Neff, Cluster Solution Engineer, will introduce CLARA, Lenovo's software-defined framework for turning logical AI cluster requirements into physically viable deployments. The session will cover algorithmic topology generation and simulated-annealing floorplan optimization. These techniques minimize cable lengths while meeting strict power and space constraints, reducing design iterations and helping prevent the cascading errors common in manual design.
For more information
For more information about the Lenovo NVIDIA Vera Rubin AI Factory Platform, see the Lenovo CSR4172 V3 NVIDIA Vera Rubin NVL72 datasheet or contact your Lenovo representative.
Author
Nevin Thomas is a marketing professional on the CSP Marketing & Strategy team in the Lenovo Infrastructure Solutions Group, based in Morrisville, NC. He leads product marketing, partner co-marketing, sales enablement and content for Lenovo’s North American cloud service provider business, which serves hyperscalers, AI cloud providers and scale cloud providers.
Trademarks
Lenovo and the Lenovo logo are trademarks or registered trademarks of Lenovo in the United States, other countries, or both. A current list of Lenovo trademarks is available on the Web at https://www.lenovo.com/us/en/legal/copytrade/.
The following terms are trademarks of Lenovo in the United States, other countries, or both:
Lenovo®
Neptune®
The following terms are trademarks of other companies:
NVIDIA®, NVIDIA Spectrum®, and NVLink® are trademarks of NVIDIA Corporation.
Other company, product, or service names may be trademarks or service marks of others.
Configure and Buy
Please select a locale
Full Change History
Course Detail
Employees Only Content
The content in this document with a is only visible to employees who are logged in. Logon using your Lenovo ITcode and password via Lenovo single-signon (SSO).
The author of the document has determined that this content is classified as Lenovo Internal and should not be normally be made available to people who are not employees or contractors. This includes partners, customers, and competitors. The reasons may vary and you should reach out to the authors of the document for clarification, if needed. Be cautious about sharing this content with others as it may contain sensitive information.
Any visitor to the Lenovo Press web site who is not logged on will not be able to see this employee-only content. This content is excluded from search engine indexes and will not appear in any search results.
For all users, including logged-in employees, this employee-only content does not appear in the PDF version of this document.
This functionality is cookie based. The web site will normally remember your login state between browser sessions, however, if you clear cookies at the end of a session or work in an Incognito/Private browser window, then you will need to log in each time.
If you have any questions about this feature of the Lenovo Press web, please email David Watts at dwatts@lenovo.com.
