NVIDIA Expands Slurm to Integrate Quantum Processors into Supercomputing Workflows

·2026.11.17 01:00

Key point

The update allows QPU builders to use a single interface for all supercomputing centers via QRMI and QDMI, supporting hybrid quantum-classical apps with CUDA-Q.

1 / 2

Details

  • Slurm 26.11 introduces a community-developed orchestration layer that lets QPU builders create a single hardware interface for every supercomputing center using a simple file, eliminating the need for bespoke QPU-specific solutions.
  • The expansion supports the QRMI and QDMI interfaces and is already adopted by Alice & Bob, IBM, IQM, Pasqal, Quantinuum, and MQSC, allowing scientists to run quantum-classical applications developed with NVIDIA CUDA-Q.
  • New CUDA-Q releases include a TCBP-JOBS decoder that is 20,000x faster than the industry standard and expanded fault-tolerant primitives that enabled Fermilab to complete studies in three weeks that typically take five months.
  • At supercomputing scale, QunaSys used cuQuantum and 1,024 NVIDIA Hopper GPUs on ABCI-Q to simulate more than 300 million quantum operations, achieving chemical accuracy for a four-hydrogen-atom model.
  • The platform expansion includes CUDA-Q Pulse for GPU-accelerated qubit control and CUDA-Q Algorithms with new tools for chemistry and materials applications, alongside integrations from Quantum Motion and Altera.

This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.

Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.