In a joint announcement on Tuesday, Microsoft (MSFT) revealed it will invest billions of dollars to acquire NVIDIA (NVDA) GPUs, with plans to share this significant computing power with the European AI startup Mistral.
This collaboration builds upon a commitment made last year, where Microsoft pledged to increase its investments by constructing new data centers in Europe. The company's president, Brad Smith, stated at the time that the move was designed to assure European clients that their operations would not be disrupted by geopolitical turbulence, particularly in the context of trade disputes.
Concurrently, Microsoft is actively embracing open-source large language models like those from Mistral. This strategy aims to reduce its reliance on other AI firms such as Anthropic and OpenAI, while also lowering AI-related costs for both itself and its customers.
Microsoft did not disclose the specific financial terms or the exact value of this deal. The company indicated it will deploy new data center capacity in Europe for Mistral, enabling the startup to utilize the NVIDIA GPUs for training and running its next-generation AI models.
Microsoft will also leverage this shared GPU resource pool to provide large model services to its Azure cloud customers in Europe. Furthermore, the tech giant plans to integrate more of Mistral's models into its own Copilot intelligent assistant ecosystem.
Comments