CUDA-Q is NVIDIA's open-source platform for hybrid quantum-classical computing, providing C++ and Python support for programming systems that integrate quantum processing units, GPUs, and CPUs. The repository contains the complete source code for the CUDA-Q toolkit, including the nvq++ compiler, the CUDA-Q runtime, and integrated CPU and GPU backends designed to accelerate application development and testing. The platform enables developers to write unified programs that orchestrate quantum and classical computing resources within a single heterogeneous system.
The toolkit is classified across multiple quantum computing domains including quantum programming, quantum simulation, quantum machine learning, and QPU interoperability. It also emphasizes high-performance parallel computing and GPU acceleration, reflecting NVIDIA's focus on leveraging graphics processors alongside quantum hardware. The repository is written primarily in C++ with Python bindings, making it accessible to both systems-level developers and application scientists.
The project maintains active development with substantial community engagement. GitGenius tracking shows a median issue and pull request response latency of 0.3 hours, indicating rapid triage and feedback cycles. Across 795 tracked items, the mean response time is 1186 hours, reflecting the full lifecycle of issues from opening to resolution. The most frequently applied issue labels are stale-notified with 285 occurrences, enhancement with 180, and bug with 95, showing that the project actively manages technical debt while prioritizing feature development. The core development team includes schweitzpgi with 447 tracked events, sacpis with 327 events, and bettinaheim with 304 events, establishing a consistent group of maintainers driving the platform forward.
The repository overlaps with other significant projects through shared contributors, linking to h2oai/h2o-3 and lightning-ai/pytorch-lightning, suggesting integration points with machine learning and data science ecosystems. This cross-project involvement indicates that CUDA-Q is positioned within a broader NVIDIA software stack for AI and scientific computing.
Documentation is comprehensive, with official installation instructions available through the CUDA-Q documentation site and development packages deployed on the GitHub Container Registry. The repository provides building instructions for developers who want to compile from source or contribute modifications. The project is licensed under Apache License 2.0 and requires contributors to accept a Contributor License Agreement, establishing clear legal frameworks for community participation.
The platform explicitly welcomes contributions through multiple channels, including the Discussions tab for feature ideas and feedback, issue tracking for bug reports, and pull requests for code contributions. Security concerns and code of conduct violations can be reported directly to [email protected]. This structured approach to community engagement, combined with the rapid response times and active triage, demonstrates NVIDIA's commitment to maintaining CUDA-Q as a collaborative, community-driven quantum computing platform.
