How the University of Utah built a sovereign AI factory to accelerate breakthroughs and slash cloud costs
University of Utah cuts AI compute costs by two-thirds with on-prem sovereign factory—and the playbook scales to enterprises managing regulated data.
Why it matters
On-premises AI infrastructure is reshaping economics for regulated workloads. The University of Utah's HPE-NVIDIA sovereign factory demonstrates measurable cost and compliance wins—and a replicable model for enterprises rearchitecting away from public cloud for IP and data control.
The key facts
8 to knowOperating expense reduction of up to two-thirds vs. public cloud alternatives
Hardware: HPE Cray XD670 servers with NVIDIA Hopper GPUs
Software stack: NVIDIA AI Enterprise, HPE Morpheus, HPE OpsRamp
Hosting: DataBank high-density colocation facility
Use cases: clinical oncology genomic analysis, mental health behavioral data processing
Regional extension: RAISE (Utah Research & AI Infrastructure for a Statewide Ecosystem) initiative
Funding model: public-private-philanthropic co-investment (university, State of Utah, Huntsman Family Foundation)
Time compression: research analysis compressed from months to hours
Go to the source
CIOcio.com
Publisher excerpt: The scaling of artificial intelligence has forced IT leaders to re-evaluate infrastructure. While public clouds offer rapid deployment for general applications, they introduce steep trade-offs for highly regulated, data-intensive workloads. Issues like high latency, unpredictable operational costs,…