News Hub

OpenAI to establish global chip factories, Meta’s AI efforts ramps up

Written by Tue 23 Jan 2024

OpenAI CEO, Sam Altman, intends to utilise funds raised for a chip venture to establish a network of semiconductor manufacturing factories. Meanwhile, Meta has announced plans to integrate artificial intelligence (AI) into its products by securing NVIDIA chips. 

Bloomberg reported that Altman has talked with several large potential investors to raise sums for chip fabrication plants, citing people familiar with the matter. 

Through the project, there will be a collaboration with leading chip manufacturers and a worldwide network of chip fabrication plants will be established.  

Abu Dhabi-based G42 and Japan’s SoftBank Group are among the manufacturers who have been in discussion with Open AI’s CEO.

However, talks are still in the early stages and a list of partners and funders has not been disclosed. Intel, Taiwan Semiconductor Manufacturing Company (TSMC), and Samsung Electronics are also potential partners for OpenAI.

Sam Altman returned to OpenAI as CEO after he was fired from the company in November. In a statement on 19 November, OpenAI said it no longer has confidence in Altman’s ability to continue leading the company. Altman was briefly hired by Microsoft to lead an advanced AI research team.

Bloomberg reported Altman had been working on the chip project before his termination and resumed his efforts upon returning to his role. Altman has also discussed the chip project with Microsoft who is also reportedly interested.

Meta Expands AI Efforts 

CEO of Meta, Mark Zuckerberg, announced plans to unite the AI research and business-focused generative AI teams, intensifying efforts to integrate technology into products. 

“We are currently training our next-gen model Llama 3, and we are building massive compute infrastructure to support our future roadmap, including 350,000 H100s by the end of the year, and almost 600,000 H100s equivalents of compute if you include other GPUs,” said Zuckerberg in an Instagram post.

Reuters reported that once finished, this quantity of GPUs would position Meta’s system as one of the largest in the industry. A Meta spokesperson did not disclose GPU suppliers beyond acknowledging the use of chips from NVIDIA. The company has publicly stated intentions to incorporate AMD chips. 

In November, Amazon announced the next generation of two chip families, AWS Graviton4 and AWS Trainium2. Trainium2 promised up to four times faster training than its predecessor and can be deployed in EC2 UltraClusters of up to 100,000 chips. 

In March, NVIDIA collaborated with Oracle to provide access to DGX supercomputers, featuring as many as 32,000 NVIDIA chips accessible through a web browser.

Join Big Data & AI World

6-7 March 2024, ExCeL London

Be at the forefront of change with thousands of technologists, data specialists, and AI pioneers.

Don’t miss the biggest opportunities to advance your business into the future.

Written by Tue 23 Jan 2024

Send us a correction Send us a news tip