Nvidia and Cisco unveil new USFP optical module for AI data centers

Nvidia (NASDAQ: NVDA) and Cisco (NASDAQ: CSCO) announced plans for a new optical module standard called Ultra-Dense Small Form Factor Pluggable (USFP), designed to support 400 gigabits-per-second (Gb/s) per lane electrical signals for use in AI data center networks.

The announcement was made at the OCP conference in San Francisco on Oct. 12, 2026.

According to the companies, USFP is designed to deliver up to 12.8 terabits per second of capacity and offers four times the bandwidth density and approximately 1.3 times the power efficiency compared with 200 Gb/s-per-lane pluggable modules of the same OSFP-sized footprint.

The first planned USFP transceivers will pack 16 full-duplex lanes in an OSFP-sized footprint, with support for both switches and SuperNICs. The companies said the design allows for expansion to 32 or more lanes in the future. USFP supports power dissipation of up to 100 watts without requiring a dedicated embedded cooling circuit within the transceiver.

"AI builders need an interconnect architecture that can scale performance without locking them into a more complex operating model," said Bill Gartner, senior vice president and general manager of optical systems and optics at Cisco.

Nvidia and Cisco are working with TE Connectivity and Amphenol to develop USFP cages, connectors and modules. Additional partners are expected to join the effort. The companies also plan to create an open industry specification for USFP through a community-led multi-source project.

The USFP initiative builds on an existing collaboration between Nvidia and Cisco to enable Nvidia Spectrum-X Ethernet networking and Cisco Silicon One-based Ethernet networks to operate together.

You May Also Be Interested In

• Jefferies Downgrades Elanco Animal Health (ELAN) to Hold
• Deepseek, Huawei Working On New Ai Chip Designs - Information
• Jefferies Downgrades Zoetis Inc. (ZTS) to Hold