Accelerated Quantum Supercomputing with CUDA-Q
Contents
As hybrid quantum–classical computing continues to mature, the ability to program heterogeneous architectures has become essential. This tutorial provides a deep dive into CUDA-Q, NVIDIA’s open-source platform for heterogeneous quantum-classical computing. Participants will learn how to leverage GPU acceleration to simulate quantum circuits, develop hybrid algorithms, and prepare for execution on current and future quantum hardware.
Technical focus areas:
- Algorithmic Implementation: A detailed look at implementing the Quantum Fourier Transform (QFT), focusing on fundamental principles, GPU-accelerated simulation, and performance benchmarking.
- Practical Applications: Moving beyond theory, the tutorial covers practical applications of GPU-acceleration in quantum computing through overviews and interactive hands-on examples.
- Hybrid AI & Machine Learning: An exploration of "AI for Quantum," where participants will program a Hybrid Neural Network using CUDA-Q, demonstrating the synergy between machine learning and quantum circuits.
- Advanced Simulation Techniques: How to leverage Tensor Networks to scale quantum algorithm simulations efficiently on classical supercomputing hardware.
Takeaways:
By the conclusion of the workshop, attendees will have a practical understanding of how to utilize LRZ’s computing services to develop, optimize, and scale quantum-classical applications. This tutorial equips researchers with the tools needed to push the boundaries of quantum simulation and hybrid algorithm design using the industry-leading CUDA-Q platform.
Materials
NVIDIA_Accelerated Quantum Supercomputing_Using_CUDA-Q_Esperanza_Cuenca-Gomez.pdf