South Korean AI chip startup FuriosaAI has announced a partnership to supply its AI chip, RNGD, to enterprises using LG AI Research’s new EXAONE platform.

RNGD is optimized for large language models (LLMs). Just last week, LG unveiled EXAONE 4.0, a next-generation hybrid AI model.

The collaboration targets sectors including electronics, finance, telecommunications, and biotechnology.

FuriosaAI Declines Meta’s $800 Million Offer

Roughly three months earlier, FuriosaAI turned down Meta’s $800 million acquisition proposal.

Local media reported that the deal collapsed due to disagreements over post-acquisition strategy and structure, not pricing.

“We want to continue our mission, and I think it’s an exciting opportunity at the same time. I believe it’s a very impactful contribution, both personally and for the company, to make AI computing more sustainable,” said June Paik, CEO of FuriosaAI.

Paik did not confirm whether the company is seeking new funding following the canceled deal.

EXAONE Partnership Expands FuriosaAI’s Global Reach

According to Paik, the new partnership with LG has potential beyond South Korea.

“LG AI’s EXAONE is regarded as the leading sovereign AI model in South Korea. It won’t be used just within LG. It will be one of the main AI models used in the Korean AI ecosystem,” he said.

“We expect there will be many demands for this EXAONE, as well as for our chip solutions in South Korea, but not only in Korea… including global customers.”

Challenging Nvidia with Cost and Performance Advantages

Paik noted that LG AI’s adoption of Furiosa’s chip marks a rare public endorsement of a non-Nvidia solution by a major enterprise.

“We had to prove that our solution not only delivers strong performance but also lowers total cost of ownership,” he said.

FuriosaAI claims its RNGD chip achieved 2.25 times better inference performance than competing GPUs with LG’s EXAONE models.

Paik also said the hardware demonstrated greater energy efficiency.

“Our architecture is natively built for AI computing. We do not develop our chip for rendering or mining,” said Paik.

“We can support a wide variety of AI models efficiently. But unlike GPUs, which are still fundamentally general-purpose processors…”

FuriosaAI is based in Seoul and maintains an office in Santa Clara. The company operates with a global team of 15 employees.

PHOTO: UNSPLASH

This article was created with AI assistance.