{"resourceId":"kit-horeka2-nhr-ai-hpc-20260925","versions":[{"version":"external-141b06f08e475be9d0aba4a671d4459560a81fbb03a2ed0b04461b056e88de6e","resource":{"id":"kit-horeka2-nhr-ai-hpc-20260925","title":"KIT inaugurates HoreKa 2: >42k cores, 300 AI accelerators, ~33 PFLOPS; doubles sim/AI compute (€17M)","organization":"Karlsruhe Institute of Technology (KIT) / NHR@KIT Scientific Computing Center","sector":"Higher education research computing (Germany NHR)","geography":"Germany (Karlsruhe)","publishedAt":"September 25, 2026","publicationDate":"2026-09-25","eventDate":"2026-09-25","sourceName":"KIT Presseinformation 081/2026","sourceLabel":"University press release (KIT)","sourceUrl":"https://www.kit.edu/kit/pi_2026_081_kit-nimmt-hochleistungsrechner-horeka-2-in-betrieb.php","evidenceClass":"independent-reporting","outcomeClass":"emerging","topics":["infrastructure","operating-model","accessibility-workforce"],"finding":"On 25 September 2026 KIT inaugurated HoreKa 2, a €17 million NHR-accessible supercomputer with more than 42,000 processor cores, 300 AI accelerators, and planned peak performance of about 33 PetaFLOPS. SCC director Martin Frank states the system doubles compute capacity for simulations and AI applications. Hot-water cooling and 100% renewable Campus Nord power are cited for efficiency; GPU partition supports AI workloads across earth system, engineering, materials, particle physics, and other disciplines. Phased rollout: CPU Onyx began earlier; GPU Jade (NVIDIA Blackwell) expands AI capacity.","sledRelevance":"Campus and national research-computing leaders evaluating AI/HPC upgrades need capacity, efficiency, and shared NHR-style access claims with clear limits—this is an inauguration with architecture specs, not measured queue SLOs or scientific discovery outcomes.","evidence":"KIT press release dated 25 September 2026 opened via HTTPS; core/accelerator/PFLOPS, €17M cost, doubling claim, and sustainability statements taken from the opened German page.","architectureImplications":"Hybrid CPU+GPU NHR cluster with hot-water cooling and staged GPU Blackwell partition; treat as shared national academic AI/HPC capacity beside prior HoreKa generation.","governanceImplications":"NHR alliance access and KIT SCC operational ownership imply allocation and fair-share policies; inauguration does not publish AI-agent evaluation or research-data classification rules.","securityPrivacyImplications":"Press release omits multi-tenant isolation, export controls, and research-data categories; plan security review before sensitive AI workloads.","caveats":"Institutional self-report; PFLOPS is planned peak; 'doubling' is operator claim without published baseline table; no queue-wait or energy-efficiency ranking metrics in this page; German national context limits direct U.S. transfer.","streamIds":["research"],"roles":{"sales":"Sales framing: Research-computing and VPR buyers want shared AI/HPC capacity that expands simulation and AI headroom without promising discovery breakthroughs. KIT’s HoreKa 2 inauguration supports a bounded pitch around >42k cores, 300 AI accelerators, ~33 PFLOPS planned peak, operator-stated doubling of sim/AI capacity, and hot-water/renewable efficiency narrative at €17M. Keep claims to national/campus shared capacity—not autonomous lab outcomes or guaranteed queue SLOs (those are not in this release).","engineering":"Engineering framing: Fit is a staged NHR CPU+GPU cluster (Onyx then Jade/Blackwell) with hot-water cooling and shared academic access. Integrate identity, fair-share scheduling, AI software modules, and GPU partition policies before migrating AI serving or training jobs. Prerequisites include power/cooling validation and migration guides; constraints include missing published wait-time telemetry on this release and German NHR allocation norms that will not map one-to-one onto a U.S. campus recharge model.","delivery":"Delivery framing: Commission remaining GPU partitions, publish allocation and SLO baselines, migrate pilot AI and simulation groups, and train users on right-sizing accelerators. Ownership sits with KIT SCC/NHR ops; PIs own scientific methods. Acceptance should require documented capacity availability and pilot workload success—not brochure peak FLOPS or TOP500 alone. Risks include phased GPU delays, efficiency claims without independent audit, and overclaiming scientific impact from inauguration coverage."},"roleDetails":{"sales":{"customerProblem":"Shared academic AI/HPC capacity is saturated while simulation and AI workloads compete for scarce accelerators.","stakeholders":["Research computing / SCC leadership","VPR / research administration","NHR allocation committees","Faculty AI and simulation users","Facilities / sustainability officers"],"discoveryQuestions":["What is current CPU vs GPU utilization and wait by partition?","Which AI workloads need Blackwell-class GPUs vs CPU-only?","How is national NHR vs campus priority allocated?","What efficiency or renewable-power constraints bind CapEx?"],"valueHypothesis":"A staged €17M-class CPU+GPU upgrade with hot-water cooling can double stated sim/AI capacity for national academic users if allocation and training keep the system shared.","potentialEngagement":"90-day pilot migration of 3–5 AI/simulation groups onto new GPU partition with utilization and wait baselines.","unsupportedClaims":["Guaranteed scientific discoveries","Measured queue-wait SLOs (not in this release)","U.S. campus cost parity","Autonomous research-agent readiness"]},"engineering":{"fit":"National academic hybrid HPC with staged NVIDIA Blackwell GPU partition beside CPU Onyx under NHR@KIT ops.","architectureIntegration":"Integrate Slurm/fair-share, AI modules, identity, storage, and GPU Jade alongside prior HoreKa; follow NHR migration docs for user cutover.","prerequisites":["Power and hot-water cooling validated","Allocation policy published","GPU software stack","User migration training","Baseline utilization metrics"],"constraints":["Planned peak PFLOPS not sustained application performance","Phased GPU rollout","German NHR access rules","No agent eval harness specified"],"security":"Define multi-tenant isolation, data classification, and export-control paths before hosting regulated research on shared AI nodes.","proofOfValue":["Partition availability dates","GPU/CPU utilization","Pilot job success rates","Independent efficiency measurement if claimed"]},"delivery":{"work":"Finish GPU partition rollout, migrate pilots, publish service catalog and allocation SLOs, train consultants.","dependencies":["Facilities cooling/power","Network/storage","NHR identity/allocations","PI pilot groups"],"ownership":"KIT SCC/NHR owns platform ops; PIs own science methods; ministry/NHR own national access policy.","skillsAdoption":"Train users on GPU right-sizing, module stacks, and fair-share; train ops on Blackwell partition incidents.","governanceCheckpoints":["Allocation policy sign-off","Security/data-class review","Pilot review at 30/90 days"],"acceptanceCriteria":["GPU partition generally available","Pilot groups productive","Published capacity vs prior system baseline"],"risks":["Phased delays leave AI users on legacy GPUs","Efficiency claims unvalidated","Oversubscription recreates queues","Transferability limits for U.S. campuses"]}},"retrievedAt":"2026-09-26T21:59:30-05:00","enrichedAt":"2026-09-26T22:08:00-05:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"NHR-wide access plus KIT training expands accelerator availability beyond single-lab GPU islands; language and allocation literacy remain barriers for non-HPC specialists.","procurementImplications":"Justify shared CapEx with capacity doubling and efficiency design, not brochure FLOPS alone; include cooling and renewable-power OpEx in TCO.","operatingModelImplications":"SCC as shared service for national academic users is the operating model that converts hardware into research access; inauguration alone does not define AI-agent governance.","sourceVerification":{"openedUrl":"https://www.kit.edu/kit/pi_2026_081_kit-nimmt-hochleistungsrechner-horeka-2-in-betrieb.php","referenceExcerpt":"Mit mehr als 42 000 Prozessorkernen, 300 KI-Beschleunigern und einer geplanten Spitzenleistung von rund 33 PetaFLOPS","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}