Microsoft's custom AI chip, codenamed Athena or Braga, has hit a significant delay, pushing its mass production to 2026, at least six months behind schedule. This delay gives Nvidia more time to solidify its market dominance in the AI chip market. The setback stems from design modifications, staffing shortages, and high employee turnover, which have repeatedly delayed the project timeline.
The delay means Microsoft will continue to rely heavily on Nvidia's high-performance GPUs, potentially driving up costs and limiting its ability to differentiate its Azure cloud services. Nvidia's Blackwell chips, with their superior performance and energy efficiency, will likely maintain their lead in the AI infrastructure market.
The delay is attributed to last-minute design changes requested by OpenAI, a key partner, which introduced instability during simulations and set the project back several months. Staffing issues, including high turnover rates, have further complicated the development process.
When released, Braga's performance is expected to fall short of Nvidia's Blackwell chip, potentially widening the performance gap between the two companies. Amazon and Google are also developing custom AI chips, intensifying competition in the data center AI market. Microsoft's delay may give these competitors an opportunity to gain ground.
The outcome of Microsoft's custom AI chip development will be closely watched, as it could have significant implications for the AI chip market and the company's cloud services offerings.
