DigestAI news desk

Cut through the AI noise.

Hardware & Compute2 min read

Nvidia and Cisco launch 400Gb/s USFP transceiver for AI data centers

Nvidia and Cisco introduced the Ultra-Dense Small Form Factor Pluggable (USFP) transceiver, a next-gen optical interface for data centers. The module supports 400Gb/s per lane, with a total capacity of up to 12.8Tb/s—four times faster than 200Gb/s interfaces—while improving power efficiency by 1.3x. It is designed for AI workloads, scaling via 16 full-duplex lanes in an octal form factor, with…

1 source

Key points

  • USFP supports **400Gb/s per lane**, with a max total of **12.8Tb/s**—four times faster than 200Gb/s interfaces
  • First versions use **16 full-duplex lanes**, with future scaling to **32+ lanes** for AI workloads
  • Third-party partners like **TE Connectivity** and **Amphenol** supply modules, connectors, and cages

Cisco notes the USFP’s two-step insertion design reduces wear and improves signal integrity, handling up to 100 watts without dedicated cooling. While full 12.8Tb/s capacity remains a future goal, the first versions will support 1.6Tb/s endpoints without adding network tiers. The companies emphasize its role in AI-driven data center growth, though Cisco acknowledges it fits into a broader mix of optical architectures.

Full story from techspot.com · by Alfonso Maruccia · via Search: NVIDIAOpen source ↗

Nvidia and Cisco unveil next-gen network interface built for the AI data center boom

techspot.com · 15 October 2026

Loading the full article…

This text was published by techspot.com and written by Alfonso Maruccia. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗

Topics · follow one to build your own front page
NvidiaCiscoTE ConnectivityAmphenol

The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.

Comments

via GitHub Discussions

More in Hardware & Compute

All →

Related stories