- Big Tech is tired of trusting nvidia so they make their own chips
- Openai’s first AI GPU is almost ready and on its way to TSMC to become tape.
- If all goes well, mass production could begin in 2026
Rising costs and a worrying dependence on AI Behemoth Nvidia have led tech giants like Microsoft, Google and Meta to look at building their own artificial intelligence chips.
Openai, involved in the newly announced $ 500 billion Stargate initiative, is also reportedly developing its own AI-Hardware to give it some freedom from Team Green.
A report from Pakinomist Requires that the company is in the last stages of producing this first chip – which could have cost Openai up to $ 500 million to design – and expect to send it for manufacture at TSMC in the coming months when mass production would probably start in 2026.
Implemented in a limited scale
Openai’s chip development has been running for a while. We first reported in July 2024 that Sam Altman’s company was in discussions with Broadcom to design and build its own silicon, and recently that the AI company was closer to this becoming a reality.
The report claims, “If the initial tape-out goes smoothly, it would allow Chatgpt Maker to mass produce its first internal AI-chip and potentially test an alternative to Nvidia’s chips later this year. Openai’s plan to send its design to TSMC this year shows that start -up has made quick progress with its first design, a process that can take other chip designers years longer. “
Led by Richard Ho, who took over Openai over a year ago and previously played a key role in the development of Google’s own custom AI processors, the team that develops the chip is reportedly relatively small, consisting of only 40 engineers.
While the internal AI-chip will be able to both workout and inference tasks, Pakinomist‘Sources say it will initially be’ implemented on a limited scale and primarily to run AI models. ‘ It will also have a limited role in the company’s infrastructure. According to the news site, Openai considers its custom AI -chip as a way to improve its negotiating position with existing suppliers, including NVIDIA.
The chip is produced using TSMC’s advanced 3-nanometer process and will have a commonly used systolic array architecture, HBM and advanced network functions.
Openai and TSMC refused to comment on the report.
