Lenovo becomes the launch partner for NVIDIA's new generation Rubin platform

Shanghai Securities News China Securities Network (Reporter Liu Yihe) reports: At NVIDIA’s GTC2026 conference held locally on March 16, Lenovo Group officially became the global first-launch cooperation partner for NVIDIA Vera Rubin NVL72, delivering fully liquid-cooled, rack-level AI systems based on this platform.

Compared with its predecessor Blackwell, Vera Rubin NVL72 integrates 72 Rubin GPUs and 36 Vera CPUs, delivering breakthroughs in both performance and cost: inference throughput per watt increases by 10x, and single-Token cost drops to one-tenth of that of the prior generation.

NVIDIA CEO Huang Renxun said that artificial intelligence has entered an era of large-scale production. “Intelligence” is being generated in real time, and enterprises need systems that can support this scale. As AI agents begin to have the ability to reason, plan, and act, the next inflection point for AI will significantly boost demand for accelerated computing, software, and AI factories. Lenovo and NVIDIA are working together to provide a complete full-stack platform to support this future.

Lenovo Group Chairman and CEO Yang Yuanqing said that Lenovo and NVIDIA’s collaboration has unique advantages, enabling them to jointly help enterprises realize AI in deployed, real-world applications—covering the entire process from experimental development to enterprise-level production, and then to AI cloud super factories. As Agentic AI (agentic artificial intelligence) drives exponential growth in inference compute demand, a lower single-Token cost has become a key issue. By combining NVIDIA AI Enterprise software with Lenovo’s full-stack hybrid AI platform and services, Lenovo can help customers achieve large-scale AI applications with higher efficiency, lower single-Token costs, and a shorter time-to-production cycle.

In addition, Lenovo Group also announced a new generation of its Hybrid AI Advantage with NVIDIA solution at the conference, aiming to accelerate AI deployment, shorten first-Token time (TTFT), and deliver measurable business outcomes across personal, enterprise, and cloud environments.

Among them, Lenovo introduced a production-ready AI platform for real-time enterprise-grade inference. The solution can deliver an investment return in under six months. Compared with similar cloud IaaS (infrastructure as a service), the single-Token cost can be reduced by up to 8x, helping enterprises bring AI workloads back to on-premises deployment with higher efficiency and stronger controllability.

Lenovo Group and NVIDIA have cooperated for many years and are among the core participants in the annual GTC conference. Currently, Lenovo Group and NVIDIA are pushing AI to achieve real production delivery globally through a new generation of AI inference platforms, AI cloud super factories, and industry-specific Agentic AI solutions.