CUDA-Q Logical
NVIDIA's logical quantum computing framework provides tools for developing, evaluating and analyzing fault-tolerant quantum programs across different quantum error-correction architectures.
NVIDIA QUANTUM RESEARCH
Latest Research by NVIDIA
Explore NVIDIA's quantum computing research across fault-tolerant compilation, quantum error correction, AI-assisted decoding and accelerated quantum-classical computing.
FAULT-TOLERANT COMPUTING
QUANTUM ERROR CORRECTION
CUDA-Q
AI FOR QUANTUM
QUANTUM-GPU COMPUTING
FEATURED RESEARCH
NVIDIA Research Highlights
CUDA-Q Logical: Retargetable Compilation for Fault Tolerant Quantum Computing
An extensible compiler infrastructure for translating logical quantum programs through quantum error-correction layers, physical schedules and real-time control plans while preserving semantics, provenance and resource information.
Composing Detector Error Models
A framework for composing detector error models across quantum operations, enabling error analysis to follow the structure of larger fault-tolerant quantum programs.
Fast and Accurate AI-Based Pre-Decoders for Color Codes
NVIDIA researchers introduce AI-based pre-decoders for triangular color codes, using local neural-network processing to improve logical error rates and decoding performance.
Fast AI-Based Pre-Decoders for Surface Codes
An AI-based pre-decoding architecture performs local and parallel error correction before residual syndromes are passed to a downstream global decoder.
QCalEval: Benchmarking Vision-Language Models
QCalEval introduces a benchmark for evaluating vision-language models on quantum calibration plots across experimental scenarios involving superconducting qubits and neutral atoms.