Pasqal Integrates Quantum Processors with Slurm HRES

Pasqal has added native Slurm HRES support so its neutral‑atom QPUs can be scheduled like any other HPC resource.

Pasqal Integrates Quantum Processors with Slurm HRES

Pasqal’s QPUs Join the Slurm HPC Ecosystem

Pasqal announced that its neutral‑atom quantum processors can now be managed through Slurm’s Hierarchical Resources (HRES) feature. HRES lets Slurm treat a quantum processing unit (QPU) as a first‑class resource, checking availability, allocating it to jobs, and releasing it automatically—just like CPUs or GPUs. The integration has been validated in a simulated on‑premises QPU environment and in Pasqal’s cloud offering, and work is under way at Italy’s CINECA supercomputing centre to connect the Pasqal QPU to the Leonardo system.

How the Integration Works

  • QRMI‑based interface – Pasqal’s Quantum Resource Management Interface, originally co‑developed with IBM, RPI and the STFC Hartree Centre, provides a vendor‑agnostic way for Slurm to query and control QPUs.
  • Native HRES support – With Slurm 26.11’s new hierarchical‑resource model, a QPU appears as a distinct resource type, enabling the scheduler to enforce exclusive access, prioritize jobs, and share the device across users without custom plugins.
  • Hybrid workflows – Users can now launch jobs that combine CPUs, GPUs, and QPUs in a single Slurm submission script, simplifying the development of quantum‑classical algorithms.
  • Open‑source tooling – Pasqal supplies middleware (Warden, QRMI) and software libraries such as Pulser, QoolQit, and a QUBO solver, while NVIDIA’s CUDA‑Q adds emulation support for Pasqal devices.

Why This Matters for HPC

Supercomputing centres rely on Slurm for resource allocation across thousands of nodes. Until now, quantum workloads required separate submission tools, creating friction for scientists who wanted to experiment with quantum acceleration. By embedding QPU management directly into Slurm, Pasqal reduces integration overhead, lowers the risk of incompatibilities, and paves the way for routine hybrid quantum‑classical experiments in production environments.

Outlook

The collaboration with NVIDIA and the ongoing deployment at CINECA illustrate a broader industry push to make quantum hardware a consumable service within existing HPC stacks. As more centres adopt Slurm‑based quantum resource management, we can expect a growing ecosystem of tools, benchmarks, and best‑practice guides that will help researchers evaluate where quantum processors add real value.

Pasqal’s integration does not claim any performance breakthrough; it is a practical step toward smoother operational use of quantum computers in high‑performance computing environments.

Source

Read original source

Why we picked this

Accepted from curated QuantumUnzipped source (relevance scoring disabled).

← Back to all articles