FuriosaAI Turns Down Meta's $800M Bid To Go Rogue With RNGD, As AI Chip War Between Nvidia, Google, And Meta Heats Up

South Korean AI hardware startup FuriosaAI has made waves in the global semiconductor industry after reportedly turning down an $800 million acquisition offer from Meta META.

The decision shows the startup's ambition to remain independent and compete directly in the rapidly evolving AI chip market dominated by NVIDIA NVDA and increasingly encroached on by tech giants like Alphabet GOOGL GOOG)) and Microsoft MSFT.

Don't Miss:

Hasbro, MGM, and Skechers trust this AI marketing firm — invest pre-IPO from $0.55 per share now.
Maker of the $60,000 foldable home has 3 factory buildings, 600+ houses built, and big plans to solve housing — this is your last chance to become an investor for $0.80 per share.

Meta's $800 Million Power Play Rejected

Meta initially called FuriosaAI after months of evaluating potential acquisition targets earlier this year. While Meta’s offer was around $300 million more than FuriosaAI's estimated market value, negotiations collapsed over differences of opinion about the company's future direction and post-acquisition structure, according to a report by Maeil Business Newspaper.

“Since October of last year, Meta had been looking at several AI semiconductor companies in the U.S. and Israel, and finally made FuriosaAI a viable acquisition target and negotiated it since the beginning of the year," an insider reportedly told Maeil Business Newspaper.

Inside South Korea's AI Chip Trailblazer

FuriosaAI, founded in 2017 by CEO June Paik, specializes in designing AI inference chips for high-performance applications as a power-efficient alternative to power-hungry GPUs. The company has about 140 employees, with more than 90% of them working in the development department. The team also includes veterans from Qualcomm QCOM, Samsung, and Google.

Trending: Deloitte's fastest-growing software company partners with Amazon, Walmart & Target – Many are rushing to grab 4,000 of its pre-IPO shares for just $0.26/share!

The startup's main product, RNGD (pronounced “Renegade”), is already getting serious hardware buzz. Launched at Hot Chips 2024, the RNGD chip is built on TSMC’s 5nm process and features dual HBM3 memory, achieving roughly twice the efficiency of traditional GPUs at just one-fourth of the power. In a world where energy-hungry AI models are growing fast, that's a pitch with global reach.

FuriosaAI has already found some notable partners, and production of RNGD is anticipated to start in the latter part of this year. According to Maeil Business Newspaper, LG AI Research and Saudi Aramco are two of the companies that have signed up to test RNGD's capabilities.

FuriosaAI has raised about $52 million in funding by now, allowing it to develop the exceptional chips and laying the groundwork in place for mass production. While this figure is nothing compared to the tens of billions raised by some of the Silicon Valley dynamos, FuriosaAI’s recent moves have cemented its place on the map.

Independent in a Chip War Era

The startup's determination to remain independent could signal bigger changes in the AI hardware ecosystem. Like the other hyperscalers, Meta is trying to reduce dependence on Nvidia's dominant chips by building or buying alternative solutions. But with specialized AI inference hardware in rising demand, startups such as FuriosaAI are increasingly unwilling to give up autonomy since they’re powered by next-generation innovation and deep technical know-how.

“Nvidia's dominance cannot last forever. There is a strong push to establish an alternative semiconductor ecosystem," CEO Paik said in an interview with The Korea Economic Daily in October. “Nvidia’s AI accelerators such as the H100 graphic processing units, sought after by companies, are in short supply and too expensive.”

FuriosaAI's decision to remain solo could turn out to be prescient. While Meta and other players are scrambling to build out vertically integrated AI stacks, the startup is going its own way, potentially redefining what it means to be a leader in AI hardware post-GPU.

Read Next: