New Data Center Performance Metrics Needed, says Nvidia


The widely used PUE metric is insufficient in the generative AI era.

Nvidia says new measures are needed to measure data center efficiency. “Supercomputer and data center operators lack measures of useful work per unit of energy to gauge their progress toward sustainable computing,” says Jeremy Rodriguez in a recent blog post.

“Today’s most widely used gauge – power usage effectiveness (PUE) – compares the total energy a facility consumes to the amount its computing infrastructure uses,” he says. Since this metric was first published by Green Grid, it has served data centers well and will continue to be useful. “But it’s insufficient in today’s generative AI era, when workloads and the systems running them have changed dramatically.”

“Modern data center metrics should focus on energy, what the engineering community knows as kilowatt-hours or joules. The key is how much useful work they do with this energy,” Rodriguez says.

Read more at Nvidia.



Related content

comments powered by Disqus
Subscribe to our ADMIN Newsletters
Subscribe to our Linux Newsletters
Find Linux and Open Source Jobs

Support Our Work

ADMIN content is made possible with support from readers like you. Please consider contributing when you've found an article to be beneficial.

Learn More”>


		<div class=