NVIDIA has announced a significant expansion of Slurm, the widely used workload manager for high-performance computing (HPC), to support quantum processors. This development aims to streamline the integration of quantum computing into supercomputing workflows, enabling researchers to leverage hybrid quantum-classical systems more effectively.
Addressing Integration Challenges
Over the past five years, quantum processor deployments at supercomputing centers have increased nearly sevenfold, with approximately 30 centers globally now deploying over 50 quantum processors. Many of these centers rely on Slurm to manage their scientific workloads. By extending Slurm's capabilities to support quantum processors, NVIDIA aims to eliminate the need for bespoke software solutions for each new quantum processor or supercomputing center.
The new release builds on existing quantum-HPC interfaces, such as the Quantum Resource Management Interface (QRMI) and Quantum Device Management Interface (QDMI). These interfaces allow quantum processing unit (QPU) builders to create a unified integration layer for all supercomputing centers. This approach simplifies the deployment process for both HPC centers and QPU developers, fostering broader adoption of quantum technologies.
Advancing Hybrid Quantum-Classical Computing
The Slurm expansion is part of NVIDIA's broader CUDA-Q platform for hybrid quantum-classical computing. CUDA-Q provides tools for integrating CPUs, GPUs, and quantum processors into unified workflows. Key advancements include:
- CUDA-Q QEC: A new quantum error correction framework with a decoder server that is 20,000 times faster than the industry standard.
- CUDA-Q Algorithms: Reusable components for quantum application development, particularly in chemistry and materials science.
- CUDA-Q Pulse: Tools for shaping qubit control pulses, with GPU acceleration for enhanced performance.
These tools enable researchers to prototype and test quantum error correction methods, design quantum algorithms, and control qubit hardware more efficiently. For example, Fermilab used CUDA-Q to complete studies in three weeks that would have otherwise taken five months.
Adoption and Ecosystem Growth
The Slurm expansion has already been adopted by leading quantum computing companies, including IBM, Pasqal, Quantinuum, and Alice & Bob. Pasqal, for instance, has integrated CUDA-Q with its QRMI runtime, allowing its neutral-atom quantum processors to function as native accelerators within Slurm-based HPC workflows.
Additionally, NVIDIA's open platform has facilitated groundbreaking research. QunaSys used NVIDIA Hopper GPUs to simulate over 300 million quantum operations, achieving chemical accuracy for a four-hydrogen-atom model. NVIDIA researchers also leveraged the platform to simulate 104 qubits under realistic noise conditions, advancing fault-tolerant quantum designs.
Looking Ahead
Slurm 26.11, the latest version with expanded quantum support, is set to launch later this month. By integrating quantum processors into familiar HPC workflows, NVIDIA's enhancements to Slurm and CUDA-Q are poised to accelerate scientific discovery and drive the adoption of hybrid quantum-classical computing.

