{"resourceId":"cra-blueprint-state-regional-ai-computing-infrastructure-20261029","versions":[{"version":"external-2905dc78745c0d26fbce8f3f91603306fdddb12ab64922d3846b05994aa00d77","resource":{"id":"cra-blueprint-state-regional-ai-computing-infrastructure-20261029","title":"CRA blueprint on state AI computing (FL, PA, UT, NY): 'hardware without people fails', governance takes longer than technology, and none of the four states is done","organization":"Computing Research Association (authors from University of Florida, Simons Foundation, University of Utah, Pittsburgh Supercomputing Center, CRA)","sector":"Higher education research computing / state AI infrastructure","geography":"United States (Florida, Pennsylvania, Utah, New York)","publishedAt":"October 2026 (released with a CRA-I post dated 2026-10-29)","publicationDate":"2026-10-29","eventDate":null,"sourceName":"Computing Research Association","sourceLabel":"From Vision to Practice: A Blueprint for State and Regional AI Computing Infrastructure and Educational Access (PDF)","sourceUrl":"https://cra.org/industry/wp-content/uploads/sites/9/2026/10/From-Vision-to-Practice-A-Blueprint-for-State-and-Regional-AI-Computing-Infrastructure-and-Educational-Access.pdf","evidenceClass":"public-sector-association","outcomeClass":"mixed","topics":["infrastructure","governance-procurement","accessibility-workforce","operating-model"],"finding":"The report, building on NVIDIA co-founder Chris Malachowsky's 50-state AI computing strategy, profiles four state efforts as of August 2026. Florida's UF program, seeded by a Malachowsky/NVIDIA gift, is the most mature; the report says every capability increase was paired with staffing and training, HITRUST auditors found 87 gaps across 288 controls that UF closed in six weeks only because of five years of NIST 800-53 work, at about $170K a year, and new institutions should plan for about five years of runway. It notes UF is effectively part of NVIDIA's sales motion to other universities. Pennsylvania's director says centralized infrastructure does not exist yet, Utah's system is months old, and New York's permanent facility is due in 2027 or 2028.","sledRelevance":"Directly relevant to governors, state higher-education systems and research universities considering NVIDIA-based state AI computing hubs. It gives practitioner lessons on staffing, compliance, ownership and funding, and is candid about vendor relationships and how early most states are.","evidence":"Practitioner report by program leaders, supported by an NSF grant and a CRA/NSF summit; descriptive case studies, not an independent evaluation; some financial-return claims are the programs' own.","architectureImplications":"Centralized shared machines with local first-line support; compliance frameworks (HITRUST, NIST 800-53) need to be designed in for sensitive research data.","governanceImplications":"Decide ownership and governance early; expect governance to take longer than deploying hardware; matching commitments signal institutional buy-in.","securityPrivacyImplications":"Supporting health and other sensitive research requires audited controls; UF's HITRUST experience shows the cost and lead time.","caveats":"Authors lead the programs described; the framing builds on an NVIDIA co-founder's strategy; status as of August 2026; outcomes such as return-on-investment claims are not independently verified.","streamIds":["nvidia"],"roles":{"sales":"Use this report with state higher-education leaders and research vice presidents as the realistic view of a state AI computing hub. Its central message is that hardware without people fails, and that governance takes longer than technology. Position engagements around readiness, staffing and governance design rather than GPU counts. Be transparent that the report builds on an NVIDIA co-founder's strategy and that UF serves as a reference customer in NVIDIA's sales motion.","engineering":"Engineering teams planning a shared state or consortium cluster should budget for staff alongside hardware and design for compliance from the start, since UF's HITRUST review found 87 gaps across 288 controls. Plan central infrastructure with local first-line support at member campuses. Expect early hardware to need production hardening; UF worked through firmware and fan failures with NVIDIA engineers. Size identity, allocation and data-transfer services for many institutions, not one.","delivery":"Delivery should treat a state AI hub as a five-year program. Settle ownership and governance before procurement, secure a matching institutional commitment, and fund staff from recurring sources because support costs cannot realistically be recovered through fees. Track milestones honestly: three of the four states profiled are still early. Report progress as active users, supported institutions and compliance status rather than installed GPUs alone."},"roleDetails":{"sales":{"customerProblem":"States and university systems want shared AI computing but underestimate staffing, governance and compliance needs.","stakeholders":["Governor's office","State higher-education system","Research VP","Research computing director"],"discoveryQuestions":["Who will own and govern the shared resource?","How will staff be funded after the initial grant?","Which compliance frameworks are required?"],"valueHypothesis":"Planning for people and governance first improves the chance that a hub is used and sustained.","potentialEngagement":"Readiness assessment and governance design for a state or consortium AI computing hub.","unsupportedClaims":["The programs' own return-on-investment claims","That the Florida model transfers to every state"]},"engineering":{"fit":"Fits state and consortium research computing initiatives.","architectureIntegration":"Central shared cluster with federated identity, allocation and local first-line support.","prerequisites":["Compliance baseline such as NIST 800-53","Staffing plan","Identity federation across institutions"],"constraints":["Early hardware may need production hardening","Five-year runway for institutions starting from scratch"],"security":"Design HITRUST or equivalent controls in for sensitive research data.","proofOfValue":["Proposed: compliance gap assessment before procurement","Proposed: pilot allocation service for two member institutions"]},"delivery":{"work":"Run a state AI hub as a multi-year program with governance, staffing and compliance milestones.","dependencies":["State funding","Member institutions","Vendor partners"],"ownership":"A named governing body with a host institution.","skillsAdoption":"Train local support staff at each member campus.","governanceCheckpoints":["Ownership decision before procurement","Annual compliance review"],"acceptanceCriteria":["Proposed: active users across member institutions","Proposed: recurring staff funding secured"],"risks":["Hardware purchased before governance is settled","Vendor reference-customer obligations"]}},"retrievedAt":"2026-10-31T02:55:33Z","enrichedAt":"2026-10-31T03:02:39Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Staff costs cannot realistically be cost-recovered, per the report; sustained funding for people is required for statewide access.","procurementImplications":"Vendor partnerships may bring early hardware and free software in exchange for reference-customer status; states should weigh that trade-off explicitly.","operatingModelImplications":"Statewide models need central infrastructure plus local support at each campus, with five-year planning horizons.","sourceVerification":{"openedUrl":"https://cra.org/industry/wp-content/uploads/sites/9/2026/10/From-Vision-to-Practice-A-Blueprint-for-State-and-Regional-AI-Computing-Infrastructure-and-Educational-Access.pdf","referenceExcerpt":"Technically and operationally, hardware without people fails: every capability increase was paired with staffing and training investment","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}