Fermilab says it has reduced the development of a fault-tolerant architecture from five months to three weeks thanks to this open source extension of the CUDA-Q platform.
Nvidia announced on September 14, 2026 the extension of its CUDA-Q open source platform with CUDA-Q Logical, an orchestration layer designed to design and test applications for fault-tolerant quantum computers. The idea is to provide a programmable and verifiable approach where discipline remains largely experimental.
The target bottleneck is co-design: today, high-level logical algorithm, error correction code, real-time decoder and quantum processor micro-architecture are developing separately. CUDA-Q Logical offers an extensible and retargetable compilation and orchestration frame that combines these four floors into a single chain.
Nvidia puts forward a result obtained at Fermilab, where the development of a fault-tolerant architecture would have gone from five months to three weeks, an acceleration of a factor seven. The platform also hosts QUOPS, a test set designed by Sandia National Laboratories to measure the extent to which fault-tolerant quantum hardware is ready for concrete applications. The target areas are drug discovery, financial modelling and material science.
Read the full article on the original source
Go to the source →