Nvidia Expands CUDA-Q Logical Quantum Platform
Analysis based on 11 articles · First reported Sep 14, 2026 · Last updated Sep 15, 2026
The launch strengthens Nvidia position as the classical computing backbone for quantum development, which could support long-term demand for its accelerated computing products even though near-term revenue impact is limited. Quantum-focused partners and hardware makers gain a common open platform, potentially accelerating commercialization timelines and increasing competitive pressure across the quantum computing ecosystem.
Nvidia announced an expansion of its open source CUDA-Q platform with CUDA-Q Logical, an orchestration layer for developing and evaluating applications for fault-tolerant quantum computers. The layer lets researchers design, coordinate and verify components such as quantum algorithms, error-correction codes, hardware architectures and QPU components, switching between configurations to find optimal system setups for logical qubits. Fault-tolerant processors using logical qubits are seen as necessary for practical applications in drug discovery, financial modeling and materials development. Early users include United States — SLAC National Accelerator Laboratory, Infleqtion, IQM, QCDesign, Quantum Motion and United States — Sandia National Laboratories. Iceberg Quantum used the platform to model a fault-tolerant architecture for Diraq qubits, reportedly showing 1,000 logical qubits could be created with about 150,000 physical qubits, roughly ten times fewer than Diraq previous estimates. Fermilab reportedly cut fault-tolerant algorithm development from five months to three weeks, a 7x speedup. United States — Sandia National Laboratories developed QUOPS, an independent cross-platform benchmark for progress toward utility-scale quantum computing, with initial results for QPUs from Alphabet Inc., IBM and Quantinuum and a reference implementation available in CUDA-Q. Nvidia also highlighted adoption of its broader quantum stack, including Ising AI models, NVQLink and cuQuantum, by Diraq, Anyon Computing, Quandela, Quantum Machines, BlueQubit, Qedma Quantum Computing, QCentroid, IonQ, Mitre Corporation, Phasecraft, UCLA and Caltech. CUDA-Q Logical is available through GitHub.
Set up alerts, explore entity relationships, search across thousands of events, and build custom intelligence feeds.
Open Dashboard