Nvidia Enriches Its CUDA-Q Quantum Platform with the CUDA-Q Logical Layer
Nvidia has unveiled an extension to its open source CUDA-Q platform with the addition of an orchestration layer called CUDA-Q Logical, designed for the design and testing of applications for fault-tolerant quantum computing. This announcement is accompanied by the integration of a measurement tool developed by Sandia National Laboratories.
An Orchestration Layer for Logical Qubit Quantum Computing
CUDA-Q Logical presents itself as an orchestration layer offering a programmable and verifiable approach for developing applications for fault-tolerant quantum computers. According to Nvidia, these processors equipped with logical qubits aim to overcome errors inherent to physical qubits, in order to execute computations of greater scale for uses such as drug discovery, financial modeling, and materials development.
The platform enables researchers to design and orchestrate the various required components (algorithm, error correction code, hardware architecture) and to switch between multiple options to identify system configurations.
Timothy Costa, vice president and quantum general manager at Nvidia, stated that the addition of CUDA-Q Logical brings "power and flexibility to explore fully integrated and co-optimized systems, regardless of qubit type and architecture".
Adoption by Laboratories and Integration of the QUOPS Benchmark
CUDA-Q Logical is already being used by quantum processor manufacturers and laboratories, including Fermi National Accelerator Laboratory, Infleqtion, IQM Quantum Computers, QCDesign, Quantum Motion, and Sandia National Laboratories. According to Nvidia, researchers at Fermilab have reduced the development of a fault-tolerant architecture from five months to three weeks.
Furthermore, QUOPS, an independent measurement tool developed by Sandia National Laboratories to assess the progress of quantum systems toward large-scale applications, is now available in CUDA-Q. Sandia has shared initial results in a preprint preceding the IEEE Quantum Week, focusing on processors from Google, IBM, and Quantinuum.
Nvidia specified that CUDA-Q Logical is available via GitHub, as is the QUOPS repository.
