{"resourceId":"uw-tillicum-dell-xe9680-h200-cluster-redesign-case-study-20261021","versions":[{"version":"external-42fd38d277e5666b80245cb329f1f4f907c5d37bf7454d396338c439d8614708","resource":{"id":"uw-tillicum-dell-xe9680-h200-cluster-redesign-case-study-20261021","title":"University of Washington Tillicum: 24 Dell XE9680 servers with 192 H200 GPUs delivered on a narrow public budget, per partner case study; TOP500 record shows rank slipping from 185 to 207","organization":"University of Washington (with Dell partner The Redesign Group)","sector":"Higher education / research computing","geography":"Seattle, Washington, United States","publishedAt":"2026-10-21","publicationDate":"2026-10-21","eventDate":"2025-10-15","sourceName":"The Redesign Group insights (Danny James)","sourceLabel":"Dell partner case study about a public university deployment","sourceUrl":"https://www.redesign-group.com/insights/university-of-washington-tillicum-ai-computing-case-study","evidenceClass":"vendor-claim","outcomeClass":"emerging","topics":["infrastructure","governance-procurement","operating-model"],"finding":"A case study published Oct. 21, 2026 by Dell partner The Redesign Group describes how the University of Washington built Tillicum, its GPU research platform, under a narrow public-university budget. Phase 1 (about 200 GPUs, a bridge to a planned 1,000+ GPU next-generation phase) comprised 24 Dell XE9680 8-way H200 servers, 3 PB of all-flash storage, NVIDIA NDR InfiniBand, Dell factory rack integration and white-glove installation; Redesign says it worked with Dell, NVIDIA and DataCore and used its partner status to get within budget. It claims research met utilization goals and that UW 'maintained and even moved up' to TOP500 #185 and Green500 #51. UW's own documentation confirms 24 XE9680 servers, 192 H200 GPUs and ~3 PB flash, with general availability on Oct. 15, 2025. The TOP500 record shows #185 at debut (Nov. 2025) and #207 in June 2026, so the 'moved up' claim is not supported.","sledRelevance":"One of few documented public-university AI clusters built on Dell servers with a stated budget constraint, phased roadmap and partner-led procurement; relevant to research universities and state systems planning shared GPU capacity. It also shows how partner marketing can overstate outcomes that public records can check.","evidence":"Partner marketing case study; no cost, utilization numbers, user counts or research outputs are disclosed. Independent corroboration: TOP500 lists the Dell-built system at 8.85 PFlop/s Rmax and 144 kW, rank 185 (11/2025) and 207 (06/2026); UW Research Computing documentation confirms the hardware and an Oct. 15, 2025 general-availability date.","architectureImplications":"Air-cooled 8-GPU H200 nodes with InfiniBand and shared flash as a bridge phase, with liquid-cooled Blackwell-class systems in a colocation or modular data center deferred to a later phase; factory rack integration shortened install.","governanceImplications":"Partner discounts and escalations shaped what a public university could afford; institutions should document how partner pricing was obtained and verify outcome claims against public benchmarks before reuse.","securityPrivacyImplications":"Not addressed; research clusters handling medical or restricted data need separate controls (UW runs a distinct restricted-access computing service).","caveats":"Partner-authored marketing; the deployment dates to 2025 and the case study is new (Oct. 21, 2026), which is why it is included now. Utilization and budget outcomes are unquantified. The TOP500 'moved up' claim conflicts with TOP500 records; Green500 #51 was not independently checked in this run.","streamIds":["dell","research"],"roles":{"sales":"Sales framing: Tillicum is a credible public-university reference for a phased Dell GPU cluster bought under budget pressure, confirmed by UW's own documentation. Use it to discuss bridge capacity before a larger liquid-cooled phase. Do not repeat the partner's claim that the system moved up the TOP500; the public record shows it slipping from 185 to 207. Ask how research computing recovers costs, who governs allocation across departments, and whether cooperative contracts beat partner-negotiated pricing.","engineering":"Engineering framing: the documented design is 24 air-cooled Dell XE9680 nodes with 192 H200 GPUs, NDR InfiniBand and about 3 PB of shared flash, delivering 8.85 PFlop/s Linpack at 144 kW per TOP500. It is a bridge phase; the next phase assumes liquid cooling in a colocation or modular site. Engineers should check facility power and cooling headroom, storage throughput for AI workloads, scheduler and container support, and whether the phase-two path can reuse networking and storage.","delivery":"Delivery framing: the UW timeline ran from deployment in August 2025 through early access to general availability on Oct. 15, 2025, aided by factory rack integration. A comparable project should plan early-access testing, a scheduler and allocation policy, user onboarding and documentation before general availability. The case study claims utilization goals were met but gives no figures, so delivery teams should define measurable utilization and queue-time targets up front."},"roleDetails":{"sales":{"customerProblem":"Public research universities need shared GPU capacity for AI research quickly, within constrained budgets and existing data center limits.","stakeholders":["VP for research / research computing director","Central IT/CIO","Procurement office","Faculty governance for allocation","Facilities"],"discoveryQuestions":["Is a bridge phase needed before a larger liquid-cooled system?","What power and cooling headroom exists in current data centers?","How will GPU time be allocated and charged across departments?","Which contract vehicle gives the best verifiable price?"],"valueHypothesis":"A phased, factory-integrated Dell GPU cluster may let a university add research AI capacity quickly; cost and utilization benefits are not disclosed.","potentialEngagement":"A bounded capacity-planning and architecture review for a phase-one research GPU cluster.","unsupportedClaims":["That the system moved up the TOP500","Any specific utilization rate or cost savings","That partner pricing will be available to other institutions"]},"engineering":{"fit":"Good fit for research institutions with air-cooled data center space and a near-term GPU gap.","architectureIntegration":"Dell XE9680 H200 nodes with NDR InfiniBand and shared flash storage under a Slurm-style scheduler with container support.","prerequisites":["Data center power and cooling capacity","Allocation and cost-recovery policy","Storage sized for AI datasets"],"constraints":["144 kW Linpack power draw for the cluster","Next phase may require liquid cooling or colocation","Restricted data requires separate environments"],"security":"Not described; require separate controls or environments for regulated research data.","proofOfValue":["Benchmark representative AI training and inference jobs","Measure queue times and utilization during early access","Verify vendor benchmark claims against TOP500 or local runs"]},"delivery":{"work":"Install, integrate and open a shared GPU cluster with early access before general availability.","dependencies":["Factory rack integration","Facility readiness","Scheduler and storage configuration"],"ownership":"Central research computing (UW Research Computing in this case).","skillsAdoption":"Researchers need onboarding on scheduler, containers and GPU programming; training offered as an option.","governanceCheckpoints":["Facility readiness sign-off","Allocation policy approval","Early-access review before general availability"],"acceptanceCriteria":["Proposed: target GPU utilization and queue-time metrics defined and reported monthly","Proposed: early-access users validate key workloads","Proposed: phase-two design documented before phase-one saturation"],"risks":["Overstated vendor or partner outcome claims","Facility limits for later phases","Underused capacity without allocation policy"]}},"retrievedAt":"2026-11-01T03:07:03Z","enrichedAt":"2026-11-01T03:11:23Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Targets researchers across all UW departments and teaching use; the case study lists end-user training as an optional service.","procurementImplications":"Shows the role of a reseller's top-tier Dell partner status and vendor escalation in fitting a public budget; buyers should compare against cooperative contract pricing.","operatingModelImplications":"Operated by UW central research computing with a phased capacity plan; Day 2 management and build-operate-transfer options were offered by the partner.","sourceVerification":{"openedUrl":"https://www.redesign-group.com/insights/university-of-washington-tillicum-ai-computing-case-study","referenceExcerpt":"The successful launch ensured they maintained and even moved up in the Top500 and Green500 rankings to #185 and #51, respectively.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}