DigestAI news desk

Cut through the AI noise.

Hardware & Computeupdated 2 min read

Nvidia claims DSX POD can cut water use up to 100% and power by up to 30%

Nvidia announced the DSX POD, a modular version of its DSX AI Factory Platform built in pre‑integrated containers. The company says the POD can shrink deployment time from six months to 14 days and is being prepared with partners including Dell Technologies, HPE, Lenovo, Foxconn and Schneider Electric for worldwide rollout in early 2027.

4 sources

Key points

  • Nvidia says DSX POD can reduce deployment time from six months to 14 days.
  • The DSX platform claims up to 100% water reduction and up to 30% power savings.
  • At 2GW scale Nvidia claims the approach can save 54 million liters of water per day.

Nvidia claims the DSX platform can reduce water consumption by up to 100 percent and power use by up to 30 percent for the same output. It connects liquid‑cooled compute with dry‑cooler‑based heat rejection that avoids evaporation, combining closed‑loop liquid cooling with outdoor dry coolers. When deployed at a 2GW scale, Nvidia says the design can save 54 million liters of water per day compared with conventional cooling‑tower systems, though some climates may need extra mechanical cooling during extreme heat.

The announcement also introduced DSX MaxLPS and DSX Flex. Ian Buck said the data centers could respond when the grid needs an extra gigawatt. MaxLPS dynamically allocates power to keep token output with 30 percent less power, while Flex lets AI data centers act as flexible grid assets, adapting workload power to utility needs, grid spikes and energy prices.

The story so far

4 episodes →
  1. Nvidia claims DSX POD can cut water use up to 100% and power by up to 30%this story
Full story from datacenterdynamics.com · via Search: NVIDIAOpen source ↗

Nvidia launches DSX POD, claims the Vera-Rubin reference architecture will make AI data centers more water and energy efficient

datacenterdynamics.com · 15 October 2026

Loading the full article…

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

Coverage and discussion

4sources
Topics · follow one to build your own front page

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