Sony gives PS Plus members extra days after major PSN outage

According to a Reuters report, OpenAI is set to begin producing its in-house AI chip next year. Sources told the outlet that OpenAI plans to finalize its design over the next few months, then send it to Taiwan Semiconductor Manufacturing Company (TSMC) for manufacturing.

By building its own chip, OpenAI won't have to use Nvidia's chips as much to train and run AI models. According to Reuters, TSMC will produce the chip using more efficient 3-nanometer technology, which will have "high-bandwidth memory" and "extensive networking capabilities."

At launch, OpenAI will deploy its in-house chip at a "limited scale" and use it mostly to run AI models, according to a Reuters report. The company also reportedly plans to develop future versions of the chip with more advanced processors and capabilities.

This comes after a Reuters report last year revealed that OpenAI was working with Broadcom to develop a custom chip. According to Reuters, OpenAI's chip design team is led by former Google TPU engineer Richard Ho and has grown from 20 to 40 people in recent months as efforts have accelerated. Tech giants like OpenAI have spent billions of dollars building out AI infrastructure and buying chips to power their data-hungry AI models. That spending doesn't look like it's going to slow down anytime soon, despite AI startup DeepSeek raising questions about whether companies really need to buy thousands of chips to power their systems.