Technology

Add as a preferred source on Google

Pasqal Links Quantum Processors to Slurm’s Hierarchical Resources

The HRES route requires Slurm 26.11 or later. Pasqal says its earlier SPANK plugin remains the production option today.

Updated 2 min read
A quantum processor sits inside a research laboratory in Palaiseau, France. (AI-generated image / EconoTimes)
A quantum processor sits inside a research laboratory in Palaiseau, France. (AI-generated image / EconoTimes)

Pasqal has connected its quantum processors to Slurm’s Hierarchical Resources (HRES) interface, allowing supercomputing centers to schedule quantum jobs alongside conventional computing resources.

The integration uses the Quantum Resource Management Interface (QRMI) and was developed by Pasqal and IBM, according to Pasqal. The company describes QRMI as vendor-agnostic and says it can support processors from both companies. Sites need Slurm 26.11 or later to use the HRES route.

Pasqal says Slurm can check processor availability, schedule jobs, control access and free resources when a job ends or is canceled. Administrators can also configure how many jobs share a quantum processing unit, or QPU.

SchedMD’s documentation says HRES entered Slurm in beta with version 25.05. NVIDIA’s technical post says Slurm 26.11 adds a plugin interface for HRES. Slurm is open-source software used to manage workloads at supercomputing centers.

Pasqal reported tests with its cloud emulator and a simulated on-premises setup. Its technical blog separately describes running a workflow on a physical QPU in Canada through Pasqal Cloud. Pasqal’s earlier QRMI SPANK plugin remains its production option, according to the company, with sites able to move to HRES as the feature matures.

Pasqal also said Italy’s CINECA supercomputing center has run jobs on its Sol QPU through Slurm using QRMI and Warden with the SPANK plugin. The company said CINECA plans to use CUDA-Q on that route; it did not describe the work as a deployment of the new HRES integration.

Pasqal Chief Technology Officer Loïc Henriet said adoption depends on integrating quantum computers into the systems supercomputing centers already use. NVIDIA Director of Quantum Product Sam Stanwyck said the project aims to help researchers apply quantum processors to demanding scientific workloads.

Related articles

View all