Openai is said to be close to launching its internal chip

Openai will remain on the right path to start producing AI chip next year, According to a report from Reuters. The outlet sources tell that Openai plans to finish its design in the next few months before sending it to the semiconductor manufacturer in Taiwan (TSMC) for manufacturing.

By making its own slide, Openai will not have to use NVIDIA chips to train and operate AI models. TSMC will produce the slide using a more efficient 3 nm technology, with “High Danger Memory” and “Wide Network Possibilities”, according to Reuters.

At launch, Openai will spread a chip at home on a “limited scale” and you will mostly use it to run artificial intelligence models, Reuters Reports. According to what was stated, the company plans to develop future versions of the chip with processors and advanced capabilities.

This follows the report last year from Reutersany Openai suggested works with Broadcom To develop a custom chip. The former Openai Richard ho chips team, from the former Google TPU engineer, is leading from 20 to 40 people in recent months with increased efforts, according to Reuters.