NVIDIA expands Slurm so supercomputers can schedule quantum processors
Original title: NVIDIA Expands Slurm to Bring Quantum Computing Into Supercomputing
Slurm 26.11, due later this month, lets QPU vendors share one interface file and reuses QRMI and QDMI.
NVIDIA said a new Slurm release will let quantum-processor builders expose hardware to supercomputing centers through a single file, reusing QRMI and QDMI instead of a custom integration for each site. Adoption is already named at Alice & Bob, IBM, IQM, Pasqal, Quantinuum, and MQSC. The same post describes CUDA-Q updates, including a TCBP-JOBS decoder NVIDIA says is 20,000x faster than the industry standard, and large simulations on ABCI-Q and JUPITER. NVIDIA says nearly 30 centers are deploying over 50 quantum processors.
Unconfirmed: The nearly sevenfold growth in QPU deployments and the 20,000x decoder claim are NVIDIA’s figures.
- Slurm 26.11 will be available later this month.
- Fermilab completed CUDA-Q studies in three weeks that NVIDIA says would normally take five months.
- QunaSys simulated more than 300 million operations on 1,024 Hopper GPUs on ABCI-Q.
- NVIDIA researchers simulated 104 qubits under noise on JUPITER.
- Vendors
- NVIDIA
From the original article
2 min read at blogs.nvidia.com
Quantum processor deployments at supercomputing centers have grown nearly sevenfold over the past five years, with nearly 30 supercomputing centers around the world now in the process of deploying over 50 quantum processors. Many of these centers use Slurm to run scientific supercomputing workloads, so expanding its support for quantum processors builds on familiar workflows […]
By Sam Stanwyck · blogs.nvidia.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.