FuriosaAI brings low-power AI chip to Europe as sovereign compute demand grows

South Korean chip startup FuriosaAI invades European datacenters

Power-efficient South Korean AI chip startup FuriosaAI has landed on European shores. On Tuesday, the chip biz revealed that it had begun fielding its RNGD -- pronounced "renegade" -- line of AI accelerators at colocation giant Equinix's LS2 datacenter in Lisbon, Portugal. Founded by June Paik and Hanjoon Kim back in 2017, before LLMs were cool, Furiosa has largely focused its attention on the South Korean domestic market, scoring wins with the likes of LG Electronics. Now the startup is looking to bring its RNGD-based inference platform to Europe, where it sees an opportunity to capitalize on growing sovereign AI compute demand. We looked at Furiosa's tensor contraction processor (TCP) architecture at this time last year. Fabbed on TSMC's 5 nm process tech, RNGD is rather modest compared to chips from Nvidia or AMD. Each PCIe card features 48 GB of HBM3, 1.5 TB/s of memory bandwidth, and, according to Furiosa, is capable of churning out 512 teraFLOPS of dense FP8 performance. The main thing going for these parts is that, compared to the competition, they're not particularly power hungry. Each card has a TDP of 180 watts. For reference, the RTX Pro 6000, Nvidia's closest competitor, offers twice the memory capacity and compute, and comparable memory bandwidth, while consuming 3.33x the power. Eight of these accelerators form Furiosa's NXT RNGD Server, a 3 kW system that boasts up to 384 GB of HBM, enough to run relatively large enterprise models like OpenAI's gpt-oss 120B, LG's Exaone 236B, or Qwen 3-30B-A3B at large context sizes and concurrency. And because the systems are air-cooled, they can be deployed in existing datacenter racks. Furiosa's foray into the European market may have more to do with brand recognition and software familiarity than offloading RNGD excess accelerator stock. As we previously reported, Furiosa is working with Broadcom on a third-generation AI accelerator. The tie-up will see the two companies adapt Furiosa's Tensor Contraction Processor tech into a multi-die system-on-package utilizing faster HBM4 or HBM4e memory. The new chip will also use Broadcom's Ethernet and PCIe switching tech to support larger scale up clusters than the eight-way systems Furiosa is already building. Furiosa is one of several companies licensing technologies from Broadcom to support its next-generation accelerators. Earlier this year, Meta unveiled its latest generation of MTIA chips built with Broadcom's help. OpenAI and Google have also disclosed chip collaborations with Broadzilla, though details remain light. Furiosa's third-gen accelerators should offer much higher performance and scalability, but their reliance on HBM4 and HBM4e memory, which are only hitting the market this year and next, means we probably won't see them in the wild anytime soon. ®

FuriosaAI brings its Nvidia-rival chips to Europe

FuriosaAI has switched on its RNGD accelerators at an Equinix datacentre in Lisbon, its first move into Europe. The pitch: a low-power alternative to Nvidia for a region chasing sovereign AI compute. A Korean chip startup wants to sell Europe a cooler, cheaper alternative to Nvidia. Its first "renegade" accelerators just went live in a Lisbon datacentre. FuriosaAI has switched on its RNGD AI accelerators in Europe. The South Korean startup is installing RNGD servers at Equinix's LS2 datacentre in Lisbon, it announced on Tuesday. The name is pronounced "renegade." The Lisbon deployment builds on a foothold Furiosa already has there. The company runs a compiler-focused R&D lab in the city, plus a new flagship office. The timing is deliberate. The reveal coincides with the RAISE Summit in Paris. It lands as European enterprises hunt for efficient AI compute they can source close to home. The efficiency pitch RNGD is built for power, not headline speed. Each accelerator uses a 5nm Tensor Contraction Processor design. It delivers 512 teraFLOPS of FP8 and holds to a strict 180-watt thermal profile. Eight of them form the NXT RNGD Server, a 3kW system. Furiosa pitches it as a dense, air-cooled inference engine that drops into standard racks. No liquid cooling or retrofits required. The efficiency gap is the selling point. The Register notes that Nvidia's nearest rival card, the RTX Pro 6000, offers twice the memory and compute. It also draws more than three times the power. "We unlock the ability for enterprises to run inference sustainably and reliably," said co-founder and chief executive June Paik. Why Europe, why now The move is as much about mindshare as sales. Europe is racing to build its own AI infrastructure and cut its reliance on American silicon. Its chip sector, meanwhile, frets about the future. A part that sips power and fits today's racks is an easier sell when energy bills are climbing. Furiosa is also building for the next round. It is working with Broadcom on a third-generation accelerator. That chip targets frontier models with a trillion or more parameters. It will use faster HBM4 memory for hyperscale inference. RNGD itself is already in mass production, made on TSMC's process with SK hynix memory. Furiosa says it has raised more than $250M to date. Why it matters Furiosa will not dethrone Nvidia. Its cards are smaller and slower. Its next chip also leans on HBM4 memory that is only now reaching the market. But that is not the bet. Some European buyers simply want a chip they can run cheaply, cool quietly, and buy outside the United States. Furiosa joins a crowded field of Nvidia challengers. They are all trying to prove there is room for more than one name in AI compute.