Delay in Microsoft’s In-House AI Chip Development Due to Concerns Over NVIDIA’s Market Competitiveness

This article does not constitute investment advice. The author holds no shares in any of the mentioned stocks.

Recent developments reported by The Information indicate that Microsoft Corporation is facing challenges with its in-house chip design efforts. As a result, the tech giant has been compelled to postpone its plans and introduce an interim chip. This decision aligns with competitive pressures as companies like Amazon and Google actively pursue custom AI processors to decrease dependence on NVIDIA’s costly offerings. While Microsoft has made strides in in-house chip development, the software-focused nature of its business has limited its options. Consequently, the company is striving to innovate and foster alternatives to NVIDIA’s technology, but design delays have necessitated a reevaluation of its strategy.

Microsoft Aims for 2027 Launch of In-House AI Chip Amidst NVIDIA Competition

A previous report from The Information noted that Microsoft’s Braga AI chip would push back its mass production timeline from 2025 to 2026. Braga was expected to succeed the Microsoft Maia chip, introduced in April 2024 and produced via TSMC’s advanced 5-nanometer process. However, the latest insights suggest that delays with Braga may also jeopardize its successors, Braga-R and Clea. Concerns surrounding these setbacks imply that the upcoming chips may not match the competitive edge of NVIDIA’s offerings, leading Microsoft to develop a compromise. The company now anticipates launching an AI chip in 2027 that will serve as a bridge between Braga and Braga-R.

Microsoft May 100 AI chip — Microsoft’s Maia 100 AI chip.

According to The Information, two sources have indicated that this new AI chip may be named Maia 280. This innovation is expected to integrate two Braga chips to enhance overall performance. Like its competitors, Microsoft is focused on achieving superior performance-per-watt metrics. Executives believe that the new chips could provide performance boosts of up to 30% compared to NVIDIA’s forthcoming 2027 models.

Presently, NVIDIA’s latest AI-focused GPUs, known as Blackwell, began shipping in 2025, and their next generation, referred to as Rubin, is on the horizon. Microsoft’s vision includes producing hundreds of thousands of in-house AI chips each year. With a growing trend among tech giants to lessen reliance on NVIDIA, companies such as Google are reportedly offering their customized AI chips for third-party computing infrastructure.

The rising demand for specialized chips has benefitted designers like Broadcom and Marvel, who are witnessing an uptick in orders. While Marvel’s shares soared 88% in 2024, they have declined by 33% year-to-date as the company grapples with unmet expectations from investors and analysts regarding AI innovations.

Although both Amazon and Google leverage their proprietary AI chips, NVIDIA continues to dominate the market. CEO Jensen Huang has pointed out that the competitive advantages of his company present challenges for custom AI chip initiatives. Despite fluctuations in NVIDIA’s stock between May and June, its shares have recently surged, affirming its market leadership driven by anticipated robust demand in the AI sector and the expanding application of GPUs in global computing infrastructures.

Source & Images