Maia 300

Microsoft announced its in-house AI accelerator, Maia 300, revealing plans to reduce reliance on NVIDIA while significantly improving power efficiency.

• •The production deployment of the third-generation in-house AI accelerator, Maia 300, is scheduled to begin in 2027.
• •The company has unveiled a strategy to scale the Maia family to multi-gigawatt levels.
• •It aims to increase generated tokens per unit of power by 40% compared to GPU-based systems.

Microsoft announced its third-generation AI chip, Maia 300, featuring TSMC's 2nm process and HBM4, with plans for a 2027 launch as a core infrastructure for its services.

• •Designed specifically for inference, it improves power efficiency by 40% compared to GPU-based systems and achieves up to seven times the processing power of the previous generation.
• •It adopts TSMC's 2nm process technology and is configured with 288GB of HBM4 memory.
• •It will be deployed at a scale of several gigawatts and positioned as the primary inference infrastructure for Microsoft's own services.

Microsoftが自社開発の第3世代AIチップ「Maia 300」を発表し、クラウドにおけるAIワークロードのコスト効率を向上させるとともに、最新の推論モデルやエージェントタスクへの対応を強化した

• •Maia 300はMicrosoftの製品やサービスにおける「主要なAI推論システム」として設計されている
• •クラウド上でAIワークロードを実行する際のコストと効率性が改善されている
• •最新の推論モデルやエージェントタスクを処理する能力が向上した

Microsoft announced the Maia 300 AI accelerator for Azure, revealing plans to shift AI inference for Microsoft 365 and Azure from NVIDIA GPUs to its own chips.

• •The primary computing devices for AI inference in Microsoft 365 and Azure will switch from NVIDIA GPUs to in-house designed chips.
• •Maia 300 is scheduled to be gradually deployed in the company's data centers starting in 2027.
• •The next-generation AI chip was announced just nine months after the previous generation.