US Edition
How AI Chip Upstart FuriosaAI Won over LG
SEOUL, SOUTH KOREA, JUL 22 – FuriosaAI declined Meta's $800 million offer to partner with LG AI Research, whose EXAONE platform uses FuriosaAI's chip delivering 2.25 times better inference performance per watt.
- FuriosaAI, a Seoul-based AI chip developer, received approval from LG AI Research to provide its RNGD accelerator for integration into LG’s Exaone AI platform scheduled for 2025.
- FuriosaAI rejected Meta's $800 million acquisition offer earlier this year due to strategic disagreements, opting to remain independent and pursue growth.
- Over a seven-month evaluation, LG verified that RNGD delivers more than twice the inference efficiency and superior energy savings compared to GPUs when running their Exaone models.
- CEO June Paik said RNGD delivers strong performance and lowers total cost of ownership while not being developed for rendering or mining tasks.
- FuriosaAI plans to expand RNGD's deployment beyond South Korea across multiple industries and expects similar partnerships in the second half of 2025.
11 Articles
11 Articles
Nvidia Chip Challenger FuriosaAI Scores First Major Customer LG
FuriosaAI Inc., the Seoul-based startup seeking to design chips to compete with Nvidia Corp., has sealed its first major contract months after rejecting an $800 million acquisition offer from Meta Platforms Inc.
FuriosaAI Now Rejects Meta, Partners With LG on AI Chips
South Korean startup FuriosaAI declined an offer to sell to Meta that was worth up to 800 million dollars, both supporting and advancing its own vision on the AI hardware market. Barely three months afterwards, the firm revealed a groundbreaking collaboration with LG AI Research to provide its customized RNGD AI chip, as one of the enterprises that can utilize the EXAONE 4.0 platform. This strategy is foundational to an increasing movement in ch…
[Chief of Staff, Sobujangban] Renegade, LG AI 'ExaOne 3.5' Fully Introduced 2.25x Improvement in Performance Per Power Compared to Existing GPUs [Digital Daily Reporter Bae Tae-yong] Domestic AI semiconductor company Furiosa AI has achieved a meaningful result by fully supplying inference accelerators to LG's large-scale language model (LLM) 'EXAONE'. This is the first case of introducing a Korean-style NPU (Neural Processing Unit) to a large-sc…
Coverage Details
Bias Distribution
- 50% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
