- Big Tech is tired of counting on Nvidia, so they make their own tokens
- The first GPU of Openai is almost ready, heading to TSMC for a banner
- If everything is going well, mass production could start by 2026
The increase in costs and a disturbing dependence on the NVIDIA giant led technology giants such as Microsoft, Google and Meta to consult their own artificial intelligence chips.
OPENAI, involved in the Stargate initiative of $ 500 billion recently announced, would also have developed its own AI equipment to give it a certain freedom of Team Green.
A report of PK Press Club Affirms that the company is in the last stages of the production of this first chip – which could have cost Openai more than $ 500 million to design – and expects to send it for manufacturing to the TSMC in the coming months, the mass production is likely to start in 2026.
Deployed on a limited scale
The development of Openai fleas has been underway for some time. We pointed out for the first time in July 2024 that the Sam Altman company was in discussion with Broadcom to design and build its own silicon and, more recently, that the Société d’Ia was getting closer to this becoming a reality.
The report says: “If the initial headband is going well, it would allow the Chatgpt manufacturer to mass produce his first internal AI chip and test an alternative to Nvidia tokens later this year. The Openai plan to send its design to TSMC this year shows that the startup made rapid progress during its first design, a process that can take other more flea designers. »»
Directed by Richard Ho, who joined Openai more than a year ago and previously played a key role in the development of personalized AI processors from Google, the team developing the chip would be relatively small, composed of only 40 engineers.
While the internal AI chip will be capable of training and inference tasks, PK Press Club“Sources indicate that it will initially be” deployed on a limited scale, and mainly for the execution of AI models “. It will also have a limited role in the company’s infrastructure. According to the media, Openai considers its personalized AI chip as a means of improving its negotiation position with existing suppliers, including NVIDIA.
The chip is produced using the advanced process at 3 nanometers from TSMC and will include a systolic network architecture commonly used, advanced networking capacities.
OPENAI and TSMC refused to comment on the report.
