Next Article
OpenAI is making its own AI chips to challenge NVIDIA
OpenAI (yep, the ChatGPT folks) is working with Broadcom to make its own AI chip, will start mass production in 2026.
The goal? Rely less on NVIDIA and AMD, and keep up with the exploding demand for smarter, faster AI hardware.
The new chip will be designed specifically for AI tasks
This new chip is designed specifically for heavy-duty AI tasks, with involvement from TSMC.
There's a huge $10 billion contract in play here, and mass production is expected to start in 2026.
OpenAI aims to boost performance and avoid supply shortages
Right now, NVIDIA pretty much dominates the AI hardware scene.
By making its own chips (with Broadcom's help), OpenAI hopes to boost performance for its models and avoid supply shortages as data centers get busier.
It's a big step toward scaling up their tech—and maybe shaking up who leads the AI race.