Choose the shape
of your workload.
Compare illustrative GPU profiles for AI applications. No capacity is available to order and these figures are not price quotes.
One sled
8 × L40S · 48 GB each
Focused inference and model experiments.
Explore profile →A whole rack
32 × L40S · four sleds
Larger models and parallel jobs.
Explore profile →A pool
64 × L40S · two example sites
Distributed application capacity.
Explore profile →