FuriosaAI is expanding availability of RNGD across Europe in collaboration with Equinix.
As an initial step, we are installing NXT RNGD servers at Equinix’s Lisbon LS2 data center, enabling European enterprises to evaluate the performance and ease of use of Furiosa’s architecture on cutting-edge models firsthand.
The announcement coincides with the RAISE Summit in Paris and follows the mass production of RNGD commencing earlier this year. At RAISE, Furiosa Co-Founder and CEO June Paik will join a panel discussion alongside leaders from CoreWeave and Supermicro, examining the critical power and scaling challenges facing global AI infrastructure, while Chief Research Officer Jeehoon Kang presents the technical advancements and upcoming roadmap of the Furiosa software stack.
The recently established Furiosa flagship office in Lisbon will support our commercial operations in Europe, along with providing technical support for customers and partners. The Lisbon office is also home to a growing team of researchers and engineers focused on compilers as well as chip and PCB design. This combination of commercial and technical functions gives the Lisbon office a unique role within the global organization.
“We are pleased to be establishing an important new distribution channel in Europe with Equinix,” said Paik. “By pairing Equinix’s infrastructure footprint designed for efficiency and sustainability with our high-performance, energy-efficient RNGD architecture, we unlock the ability for enterprises to run inference sustainably and reliably.”
Navigating Europe’s energy and grid constraints
Expanding traditional, power-hungry AI infrastructure in Europe has become increasingly complex due to severe power grid limitations and energy constraints. Scaling AI compute under these conditions requires an alternative to legacy architectures that demand massive power grid overhauls or long infrastructure lead times.
RNGD addresses these power limitations directly, offering high performance along with superior tokens-per-watt efficiency built on Furiosa’s proprietary Tensor Contraction Processor (TCP) architecture.
A better AI inference option for European enterprise customers
The Lisbon LS2 deployment delivers immediate deployment and evaluation advantages to enterprise customers:
- Immediate deployment without infrastructure upgrades: RNGD drops seamlessly into standard, air-cooled data centers. This allows enterprises to deploy high-performance compute immediately for large language models and agentic AI applications, without waiting for liquid-cooled infrastructure or new local electricity generation to come online.
- Live performance evaluation in Lisbon: As an initial step in this collaboration, Furiosa is currently installing RNGD servers in Equinix’s Lisbon LS2 data center. Working alongside Furiosa’s Lisbon office, these servers will act as live demonstration hubs.
- Firsthand model testing: European enterprises can utilize these on-site environments to evaluate the practical execution speed, ease of use, and workload performance of the RNGD architecture on cutting-edge models firsthand.
- Mature software integration: RNGD is supported by a mature SDK that eliminates the need for maintaining vast libraries of hand-tuned kernels, simplifying engineering workflows for rapid deployment.
By pairing Equinix’s premier data center footprint with Furiosa’s highly efficient hardware, enterprise teams can now scale advanced, production-ready AI applications reliably and sustainably.
Written by
The Furiosa Team
