{"resourceId":"sdsc-expanse2-nsf-award-20260915","versions":[{"version":"external-311657e349efee4cacf00e29ee3061d0601e50b9a167d2ae5258c554be940621","resource":{"id":"sdsc-expanse2-nsf-award-20260915","title":"NSF Expanse2 award funds 2027 SDSC capacity for AI/HPC research; outcomes still prospective","organization":"San Diego Supercomputer Center at UC San Diego (NSF award)","sector":"University research computing / national cyberinfrastructure","geography":"United States (California; national ACCESS/NAIRR user community)","publishedAt":"September 15, 2026, SDSC news (PR20260916)","publicationDate":"2026-09-15","eventDate":"2026-09-15","sourceName":"San Diego Supercomputer Center","sourceLabel":"University/center award announcement for NSF Expanse2","sourceUrl":"https://www.sdsc.edu/news/2026/PR20260916-Expanse2.html","evidenceClass":"vendor-claim","outcomeClass":"emerging","topics":["infrastructure","accessibility-workforce","operating-model","governance-procurement","developers-agents"],"finding":"NSF awarded $10M for Expanse2 to expand AI/HPC capacity and NAIRR-facing access, but production is planned for 2027 and no Expanse2 scientific-outcome metrics are published yet.","sledRelevance":"Directly relevant to U.S. university research computing and NAIRR-linked infrastructure planning; announcement defines capacity and education goals, not measured discovery gains.","evidence":"Published September 15, 2026: $10M NSF award (2614012) to SDSC/UC San Diego to deploy NSF Expanse2 replacing Expanse (since late 2020); AMD processors/accelerators; all-NVMe AI/ML storage plus capacity tier; Jupyter and AI/ML frameworks; EOT for universities, teaching colleges, high schools; industry discretionary allocations; production in 2027; intended NAIRR contribution. No utilization, time-to-science, or reproducibility metrics for Expanse2 reported.","architectureImplications":"Interpretation: plan heterogeneous CPU/GPU scheduling, NVMe-hot paths for AI workflows, and gateway/Jupyter access alongside batch—while keeping allocation policies distinct from scientific acceptance criteria.","governanceImplications":"Interpretation: treat the award as a capacity roadmap; require allocation transparency and preregistered service metrics before claiming AI-enabled discovery acceleration.","securityPrivacyImplications":"Interpretation: federated data ecosystem claims still need campus data-classification rules and sponsor-specific residency controls for NAIRR-linked workloads.","caveats":"Prospective system entering production in 2027. Institutional announcement, not an independent evaluation. Architecture details are high-level. Prior Expanse user counts do not prove Expanse2 outcomes.","streamIds":["research"],"roles":{"sales":"Interpretation: Engage research computing directors and VPR offices tracking NSF ACCESS/NAIRR capacity pipelines. Ask which workloads need 2027 GPU/NVMe paths, how allocations will be prioritized, and what acceptance metrics research offices require before migrating from Expanse or commercial APIs. Offer a bounded readiness assessment for one college—identity, storage classes, and allocation policy—not a discovery-speed guarantee. Value hypothesis is clearer national capacity timing. Unsupported: measured Expanse2 scientific gains or guaranteed NAIRR wait-time reductions.","engineering":"Interpretation: Fit is national research HPC with batch and science-gateway users needing AI/ML frameworks beside traditional simulation. Integrate campus IdP, Jupyter, and approved data paths to Expanse2 allocations when available. Prerequisites include allocation readiness and workflow portability tests. Constraints: 2027 production horizon and AMD/accelerator software stack maturity. Security: classify data before federated ecosystem use. Proof of value: pilot job success, queue wait, and storage I/O on a preregistered workload set—not publication counts.","delivery":"Interpretation: Research computing owns operations; PIs own scientific methods; NAIRR/ACCESS processes own eligibility. Work is porting pilots, documenting eligible use, and training for AI/ML libraries on the new stack. Dependencies: NSF award schedule and facility operators. Governance checkpoints: allocation policy approval and pre-production shadow runs. Acceptance: authenticated allocations, published scheduling rules, and logged job completion for a pilot cohort. Risks: overselling 2027 capacity and confusing access with validated science."},"roleDetails":{"sales":{"customerProblem":"Campuses need foreseeable national AI/HPC capacity for long-tail science without overstating when benefits will arrive.","stakeholders":["Research computing leadership","Vice president for research","NAIRR/ACCESS liaisons","College research facilitators"],"discoveryQuestions":["Which workloads are blocked on current Expanse or commercial GPU limits?","What metrics define successful Expanse2 onboarding?","Who prioritizes AI training vs traditional HPC queues?","Which data classes can use federated ecosystems?"],"valueHypothesis":"A timed national system can reduce ad-hoc commercial GPU sprawl if allocation and storage policies are explicit before production.","potentialEngagement":"Bounded 2027 readiness assessment for one research center: portability checklist, data classes, and allocation strategy.","unsupportedClaims":["Measured discovery acceleration from Expanse2 today","Guaranteed NAIRR wait-time cuts","Automatic compliance with every sponsor rule"]},"engineering":{"fit":"National NSF HPC resource for mixed AI/ML and traditional simulation via batch and gateways.","architectureIntegration":"Map campus workflows to Expanse2 queues, Jupyter, NVMe-hot storage, and capacity tier; retain local scratch/egress controls.","prerequisites":["Working allocation pathway","Container/module portability tests","Data classification","User training plan"],"constraints":["2027 production date","AMD/accelerator software ecosystem","National queue competition"],"security":"Enforce data-class rules on federated datasets; audit gateway tokens and shared project spaces.","proofOfValue":["Pilot job success/error rates","Queue wait vs baseline","I/O throughput on representative AI workflows"]},"delivery":{"work":"Prepare pilot ports, publish eligible-use guidance, and train first research teams before production claims.","dependencies":["SDSC/NSF schedule","Campus identity and storage teams","PI time for workload selection"],"ownership":"Research computing operates access; PIs own scientific validation.","skillsAdoption":"Short workshops on AI/ML frameworks, Jupyter vs batch, and storage tier choice.","governanceCheckpoints":["Pre-production allocation policy sign-off","Shadow-run review","30-day pilot incident review"],"acceptanceCriteria":["Pilot users run authenticated jobs","Scheduling rules published","Telemetry retained for pilot cohort"],"risks":["Premature benefit claims","Queue contention","Data misclassification on federated paths"]}},"retrievedAt":"2026-09-17T03:03:34Z","enrichedAt":"2026-09-17T03:07:50Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Interpretation: EOT across teaching colleges and high schools can widen access, but still teach allocation literacy and data classification.","procurementImplications":"Interpretation: compare national allocation opportunity cost with commercial GPU contracts using local workload telemetry before multi-year cloud commitments.","operatingModelImplications":"Interpretation: central research computing brokers ACCESS/NAIRR; colleges sponsor pilots; PIs remain accountable for scientific claims.","sourceVerification":{"openedUrl":"https://www.sdsc.edu/news/2026/PR20260916-Expanse2.html","referenceExcerpt":"When it enters production in 2027, NSF Expanse2 will be one of the newest resources available","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}