{"resourceId":"nvidia-newsroom-jupiter-juelich-104-qubit-fault-tolerant-simulation-20261019","versions":[{"version":"external-fad5153583a080104d24d0526363880b2edd4d7721f6e70fffddd419a38394df","resource":{"id":"nvidia-newsroom-jupiter-juelich-104-qubit-fault-tolerant-simulation-20261019","title":"NVIDIA newsroom: record 104-qubit fault-tolerant quantum simulation on Jülich's JUPITER exascale system using 8,192 GH200 GPUs","organization":"NVIDIA with Jülich Supercomputing Centre (Forschungszentrum Jülich)","sector":"Public research supercomputing","geography":"Germany / Europe","publishedAt":"2026-10-19","publicationDate":"2026-10-19","eventDate":"2026-10-19","sourceName":"NVIDIA Newsroom","sourceLabel":"Vendor press release","sourceUrl":"https://nvidianews.nvidia.com/news/nvidia-runs-record-simulation-of-fault-tolerant-quantum-computing-at-julich-supercomputing-centre","evidenceClass":"vendor-claim","outcomeClass":"emerging","topics":["infrastructure","developers-agents"],"finding":"NVIDIA says its researchers ran a 104-qubit simulation of magic-state cultivation, a fault-tolerant quantum computing approach, under realistic hardware noise on JUPITER, Europe's first exascale supercomputer at Jülich, running on NVIDIA GH200 superchips and scaling to 8,192 GPUs. NVIDIA calls it the largest comprehensive simulation of a fault-tolerant quantum processor to date and says it validates an approach needing 90% fewer physical qubits than previous methods. Development used a GB200 NVL72 system at NVIDIA's Accelerated Quantum Research Center, scaling tests from four GPUs to hundreds before moving to JUPITER, which NVIDIA says reduced a weeks-long effort to days. Software was cuQuantum and the CUDA-Q fault-tolerant logical layer. JSC director Kristel Michielsen is quoted. An earlier Jülich result simulated 50 qubits under ideal, noiseless conditions.","sledRelevance":"Relevant to university and national-lab research computing: it shows a publicly funded supercomputer used to simulate quantum hardware designs, and a staging pattern (small development system first, then scarce exascale time). U.S. research institutions planning quantum or HPC allocations can note the staging approach. Practical impact for most SLED buyers is limited and long-term.","evidence":"Vendor press release with a quoted public-centre director. No logical error rate, fidelity or paper is cited in the release. The 90% figure is presented without stating its source or unit; see the CTOL analysis item in this edition.","architectureImplications":"GPU supercomputer as a quantum-design simulator; staged development on a smaller GB200 system before large allocations; cuQuantum and CUDA-Q software stack.","governanceImplications":"Public supercomputer time used for vendor research; allocation and access arrangements are not described.","securityPrivacyImplications":"Not applicable to this source.","caveats":"Vendor claim; no peer-reviewed paper linked. 'Validates' is NVIDIA's characterization. Applications in drug discovery and finance are prospective.","streamIds":["nvidia"],"roles":{"sales":"Sales framing: for university and national-lab research-computing leaders, this is an example of a public supercomputer used with NVIDIA software to simulate quantum hardware designs before they are built. Interpretation: the more transferable point is the staging approach that conserved scarce exascale time. Headline claims such as 90% fewer qubits are vendor characterizations questioned by independent analysis, so do not repeat them as validated results in customer conversations.","engineering":"Engineering framing: the run used cuQuantum state-vector simulation and the CUDA-Q fault-tolerant layer at up to 8,192 GPUs on GH200 nodes, after development on a GB200 NVL72 system. Pre-sales engineers supporting research institutions can discuss this staging pattern and GPU-hour budgeting from small runs. The release reports no logical error rate, so the scientific outcome is not established by this source.","delivery":"Delivery framing: research-computing teams running similar projects need allocation proposals, staged scaling tests and reproducible software environments across development and production systems. Delivery teams can adopt the small-to-large validation sequence described here. Teams should budget GPU-hours from small runs, record the software versions used at each scale, and require that results include the scientific metric, such as a logical error rate, before they are cited outside the project. Acceptance criteria below are proposals, not observed outcomes."},"roleDetails":{"sales":{"customerProblem":"Research institutions need to justify scarce HPC allocations for emerging quantum research.","stakeholders":["University research computing director","National or state lab leadership","Quantum research faculty"],"discoveryQuestions":["Do researchers have a smaller system to stage work before large allocations?","Is quantum simulation part of your research roadmap?","How are GPU-hours budgeted?"],"valueHypothesis":"Interpretation: staged development can reduce wasted large-allocation time; NVIDIA's weeks-to-days claim is unverified.","potentialEngagement":"Research-computing allocation and staging-plan review for quantum simulation workloads.","unsupportedClaims":["That the 90% qubit saving is validated","Near-term drug-discovery or finance impact","Any logical error-rate result"]},"engineering":{"fit":"HPC and quantum-simulation reference for a specialist audience.","architectureIntegration":"cuQuantum and CUDA-Q on GH200 supercomputer nodes; GB200 NVL72 development system.","prerequisites":["Large distributed GPU memory","Specialist simulation code","Allocation on a large system"],"constraints":["Memory doubles per added encoded bit","No published paper yet"],"security":"Not applicable to this source.","proofOfValue":["Reproduce small-circuit validation before scaling","Budget GPU-hours from small runs"]},"delivery":{"work":"Plan staged scaling and manage allocation use.","dependencies":["Development cluster","Allocation award","Software stack"],"ownership":"Research-computing centre with principal investigators.","skillsAdoption":"Requires quantum-information and HPC expertise.","governanceCheckpoints":["Allocation review","Results publication before public claims"],"acceptanceCriteria":["Proposed: small-scale validation passes before large runs","Proposed: published methods and error metrics"],"risks":["Overstated results","Allocation exhausted before validation"]}},"retrievedAt":"2026-10-20T02:54:12Z","enrichedAt":"2026-10-20T03:03:02Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Requires specialist quantum and HPC skills; no training element described.","procurementImplications":"Supports budgeting smaller development systems alongside large-allocation requests; no purchase described.","operatingModelImplications":"Staging work on a smaller system before using scarce supercomputer time is a transferable allocation practice.","sourceVerification":{"openedUrl":"https://nvidianews.nvidia.com/news/nvidia-runs-record-simulation-of-fault-tolerant-quantum-computing-at-julich-supercomputing-centre","referenceExcerpt":"the 104-qubit simulation is the largest comprehensive simulation of a fault-tolerant quantum processor to date","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}