UfiSpace Debuted Liquid-Cooled AI Networking Platform
The company showcased high-capacity hardware designed to manage extreme power densities in future AI data centers.
Updated on Oct. 13, 2026 in Data Centers
Live Poll
Do you trust that new AI infrastructure technology will successfully reduce long-term data center energy consumption?
UfiSpace demonstrated a liquid-cooled 1.6T networking platform at the OCP Global Summit in San Jose. The technology aims to address the thermal constraints and power demands of next-generation artificial intelligence workloads.
Why it matters
AI clusters require advanced cooling and high-speed data transmission to prevent hardware failures caused by extreme heat and power density. These solutions are intended to ensure consistent performance for large-scale computing tasks.
The S9331-64HO3L platform features 64 x 1.6T OSFP ports and utilizes a Cooling Distribution Unit for thermal management. Other exhibited hardware includes the 102.4Tbps S9331-64HO air-cooled platform and the 51.2Tbps S9630-32HO router.
The players
UfiSpace
This company specializes in open networking solutions and disaggregated hardware for telecommunications and data center operators.
GeNopsys Technologies
This firm develops optical networking components and collaborates on integrated Ethernet and optical switching systems.
The details
The platform incorporates a Cooling Distribution Unit to validate thermal feasibility for extreme AI hardware requirements. Additionally, a collaboration with GeNopsys Technologies introduced integrated Ethernet-optical switching to eliminate O-E-O conversion for improved efficiency.
Timeline
October 12-15, 2026: OCP Global Summit held in San Jose.
2027/2028: Expected deployment of AI clusters utilizing these liquid-cooled architectures.
The Tech Race
The transition to 1.6T networking platforms marks a significant industry pivot toward higher bandwidth to support modern AI workloads. This hardware effectively replaces existing 800G systems as companies compete to provide the highest capacity for data-intensive cluster communications.
For data center operators, these technologies provide a path to scale AI infrastructure without hitting current thermal or throughput ceilings. Improved buffering on routers also helps ensure more reliable data transmission for high-demand enterprise applications.
The takeaway
Liquid cooling is becoming a standard necessity for managing the heat generated by high-density AI processing hardware. As data centers scale, companies are increasingly shifting away from traditional air-cooling methods to maintain system stability.
Further reading
Learn more about evolving infrastructure requirements at the Data Centers section.
More information
View technical specifications and updates on the UfiSpace AI networking solutions official page.
Live Poll
Do you trust that new AI infrastructure technology will successfully reduce long-term data center energy consumption?
