Microsoft Corp. MSFT and Nvidia Corp. NVDA are significantly expanding their supercomputing partnership, with both the company’s CEOs highlighting a “40x speed-up” in Azure-powered artificial intelligence infrastructure during Microsoft’s Build 2025.What Happened: Microsoft CEO Satya Nadella and Nvidia CEO Jensen Huang revealed that their collaborative AI supercomputing efforts have reached full production with the new Grace Blackwell architecture, representing dramatic performance improvements over previous generations.“We are ramping and scaling and building the largest AI supercomputer in the world in Azure,” Huang said, noting that the companies’ joint innovations across the entire technology stack have yielded “40x speed-up over Hopper” in just two years.The partnership leverages Nvidia’s new FP4 Tensor Core architecture, liquid cooling technology, and advanced NVLink capabilities that connect Grace and Blackwell processors coherently. These hardware innovations combine with new CUDA algorithms and Microsoft’s Azure infrastructure to deliver unprecedented AI performance gains.Huang emphasized the importance of regular hardware upgrades to maximize performance benefits: “When technology is moving 40 times per generation and it’s 40 times every two years, you really want to upgrade every year.”Why It Matters: Both executives highlighted how software optimization continues to benefit even older hardware generations. “Hopper on Hopper has been 40x over the last two years,” Huang noted, attributing improvements to new techniques like “in-flight batching and speculative decoding.”The alliance extends beyond AI training to accelerating diverse computing workloads. “Our two teams are working on accelerating data processing by 20x, 50x, transcoding video, image processing, models of all kinds,” Huang said.Nadella summarized the partnership’s goal as delivering more “tokens per dollar per watt” to help “the world flourish” through advanced computing technologies that neither company could achieve independently.

MSFT

NVDA

Microsoft and Nvidia are expanding their supercomputing partnership, achieving a "40x speed-up" in Azure-powered AI infrastructure. CEOs Satya Nadella and Jensen Huang announced the full production of their collaborative AI supercomputing efforts with the new Grace Blackwell architecture. This partnership utilizes Nvidia's advanced hardware and Microsoft’s Azure to enhance AI performance significantly. The alliance aims to accelerate various computing workloads and optimize software for older hardware, ultimately delivering more efficiency in computing technologies.

- Microsoft and Nvidia are expanding their supercomputing partnership, achieving a "40x speed-up" in AI infrastructure with the new Grace Blackwell architecture.  
- The collaboration utilizes Nvidia's advanced hardware and Microsoft’s Azure, enhancing performance across various computing workloads.  
- Both CEOs emphasized the importance of continuous hardware upgrades and software optimization to maximize efficiency and drive innovation in AI technologies.  

Jensen Huang, Satya Nadella Tout '40x Speed-Up' In Azure-Powered AI As Microsoft, Nvidia Deepen Supercomputing Alliance