Microsoft Turns to AWS as GitHub Faces AI Capacity Crunch

According to RuntimeWire, Microsoft has begun purchasing compute capacity from AWS to support GitHub Copilot's growing demands. The move is reportedly due to Azure's inability to procure enough NVIDIA H100 GPUs to meet inference needs. This is unusual because Microsoft typically prioritizes Azure for its services. The article suggests that the global GPU shortage has forced a temporary multi-cloud strategy. Microsoft declined to comment, but sources indicate that the arrangement is limited to specific regions. This development highlights the intense competition for AI compute resources and the challenges even hyperscalers face in scaling infrastructure.