Satya Nadella, Microsoft’s CEO, posted a video on Thursday showcasing his firm’s initial deployed massive AI system, which Nvidia refers to as an AI “factory.” He stated that this is the “first of many” Nvidia AI factories to be deployed across Microsoft Azure’s global data centers for OpenAI workloads.
Each system is a cluster of over 4,600 Nvidia GB300 rack computers, each featuring the highly sought-after Blackwell Ultra GPU chip and connected through Nvidia’s high-speed InfiniBand networking technology. (In addition to AI chips, Nvidia CEO Jensen Huang strategically dominated the InfiniBand market by acquiring Mellanox for $6.9 billion in 2019.)
Microsoft has pledged to deploy “hundreds of thousands of Blackwell Ultra GPUs” as it implements these systems globally. While the scale of these systems is impressive (and the company provided extensive technical details for hardware enthusiasts), the timing of this announcement is also significant.
This announcement follows OpenAI, its partner and well-known frenemy, securing two major data center agreements with Nvidia and AMD. By some estimates, OpenAI has committed to $1 trillion in investments to build its own data centers in 2025. CEO Sam Altman announced this week that more are planned.
Microsoft is making it clear that it already possesses the data centers—more than 300 across 34 countries—and is “uniquely positioned” to “meet the demands of frontier AI today,” according to the company. These massive AI systems can also handle the next generation of models with “hundreds of trillions of parameters,” it stated.
We anticipate further details regarding Microsoft’s efforts to support AI workloads later this month. Microsoft CTO Kevin Scott is scheduled to speak at TechCrunch Disrupt, which will take place in San Francisco from October 27 to October 29.
