Microsoft to Continue Buying AI Chips From Nvidia and AMD

Microsoft will continue to purchase artificial intelligence (AI) chips from leading manufacturers Nvidia and AMD, even after introducing its in-house AI silicon, according to CEO Satya Nadella. The remarks, made during a recent interview, highlight Microsoft's strategy to maintain robust supply lines and collaborate across the technology industry as AI applications place unprecedented pressure on chip resources.

Microsoft announced in late 2023 the development of two proprietary chips — the Maia AI Accelerator and the Cobalt CPU — as part of a broader effort to optimize its Azure cloud platform for AI workloads. These chips are designed to handle large-scale training and inference for generative AI and machine learning models, at a time when demand for such technologies is soaring worldwide.

However, Nadella clarified that introducing Microsoft’s own hardware does not mean it will cut off reliance on third-party suppliers. “We’re thrilled to have great partnerships with both Nvidia and AMD, and that will continue,” he stated. The CEO emphasized that while internal innovations like Maia and Cobalt provide strategic flexibility and drive Azure’s performance, their role will be complementary rather than exclusive.

Since the explosion of interest in generative AI—sparked by language models such as OpenAI’s GPT—the market for graphics processing units (GPUs) and specialized AI chips has dramatically tightened. Nvidia’s dominance in AI silicon, particularly with its H100 GPUs, has led to shortages and long wait times for enterprise customers worldwide. Maintaining relationships with Nvidia and AMD allows Microsoft to ensure a steady supply for its ever-expanding generative AI portfolio, which includes Copilot and its integration with services like Microsoft 365 and Azure OpenAI Service.

Nadella also pointed to the strength of the broader semiconductor ecosystem, stating, “Our goal is to have the best cloud for AI, powered by the best chips—whether they come from our own labs or from our partners.” Microsoft’s multi-pronged approach mirrors a wider trend among cloud giants and major AI players, including Google and Amazon, who have also built custom chips while maintaining longstanding partnerships with established suppliers.

This hybrid sourcing model not only buffers against industry bottlenecks but fosters innovation through competition and collaboration. For European businesses and governments increasingly investing in digital transformation and AI adoption, Microsoft’s strategy signals continued stability in cloud and AI service delivery—even as global chip competition remains fierce.

For further details, visit the original article on TechCrunch.