Pasqal Demonstrates Hybrid GPU-to-QPU Workflows Using Slurm 26.11 Hierarchical Resources, QRMI, and CUDA-Q
quantumcomputingreport.com Nov 17, 2026

Pasqal Demonstrates Hybrid GPU-to-QPU Workflows Using Slurm 26.11 Hierarchical Resources, QRMI, and CUDA-Q

AI-summarised brief · reviewed before publication

Pasqal, in collaboration with IBM and NVIDIA, showcased an end‑to‑end quantum‑HPC scheduling solution at SC26 that integrates NVIDIA’s Slurm 26.11 Hierarchical Resources (HRES) framework with the open‑source Quantum Resource Management Interface (QRMI). The Lua‑based qrmi_hres.lua plugin replaces legacy license‑tracking methods, enabling Slurm to monitor neutral‑atom QPU availability, open authenticated sessions via Pasqal’s Warden middleware, and schedule jobs across cloud and on‑premise quantum processors. Deployed on Italy’s CINECA supercomputing center using the Sol neutral‑atom processor, the system lets administrators set QPU concurrency limits and maintenance intervals directly in Slurm’s resources.yaml. A demonstrated workflow uses Pulser and CUDA‑Q 0.17 to run a GPU‑based parameter sweep before automatically launching the same task on a physical Pasqal FRESNEL_CAN1 QPU if the simulation succeeds. The integration supports automatic token handling through MUNGE authentication, ensuring secure, queued access during hardware calibration and clean release on job completion or cancellation.

💡 Why It Matters

  • · By unifying quantum and classical resource scheduling within a single Slurm environment, the solution removes operational silos and accelerates the transition from simulation to real‑hardware execution.