A recent Q2BTokyo panel brought together leading experts in quantum computing and high-performance computing (HPC) to discuss the integration of these technologies. The panelists included Yuval Boger, Chief Commercial Officer of QuEra, Ryousei Takano from AIST, Elica Kyoseva from NVIDIA, and Oktay Goktas, founder of Agnostic. Each shared their insights on the current state and future of quantum computing in HPC environments.
Ryousei Takano, from AIST, has a background in operating systems and HPC, and is currently involved in managing the world’s first quantum-dedicated large-scale HPC system, ABCIQ. Elica Kyoseva, Director of Quantum Algorithm Engineering at NVIDIA, focuses on quantum algorithms for drug discovery and healthcare. Oktay Goktas emphasized his work on creating software layers that bridge HPC, quantum, AI, and cloud technologies together.
Takano discussed the integration of quantum and classical computing infrastructures, highlighting the complexity of managing both systems. He noted the importance of low-latency communication for iterative algorithms, such as those used in quantum chemistry and optimization.
Goktas elaborated on the challenges of managing diverse infrastructures, emphasizing the need for user-friendly interfaces. He pointed out that most users are accustomed to programming in Python, which necessitates simplifying the tools required to operate these complex systems.
Hybrid algorithms, which combine classical and quantum computing, are becoming more prevalent. Kyoseva mentioned NVIDIA’s CUDA-Q product, which facilitates seamless control and job distribution across different processing units (GPU, CPU, and QPU). She stressed that even as quantum computers improve, classical computing will remain integral to HPC centers due to its role in error mitigation and algorithm development.
{{Newsletter-signup}}
The panelists discussed the growing demand for on-premises quantum computers. Takano explained that on-premises installations are crucial for applications requiring low-latency interactions between quantum and classical systems. Kyoseva added that integrating quantum computers with AI supercomputers on-premises allows for better development of digital twins and hardware optimization.
The panelists also explored what quantum vendors could learn from classical HPC practices. Kyoseva highlighted the importance of enabling domain experts to access quantum computing resources without intermediaries, similar to how NVIDIA democratized access to GPUs with CUDA.
Goktas noted the differences between the quantum and classical HPC communities, emphasizing quantum computing's accessibility to a broader range of users, including younger, less specialized individuals.
The discussion touched on the importance of building national and regional quantum computing ecosystems. These initiatives not only focus on hardware but also aim to foster communities and startups around quantum technologies. Countries like Finland prefer local installations to support their national programs and build expertise.
Integration Complexity: Managing hybrid HPC and quantum systems remains complex, with a need for simplified, user-friendly tools.
Hybrid Models: Classical and quantum computing will continue to coexist, with hybrid models playing a crucial role in algorithm development and error correction.
On-Premises Demand: Low-latency requirements and data locality are driving demand for on-premises quantum installations.
Learning from HPC: The quantum industry can learn from classical HPC's approach to democratizing access and supporting diverse user communities.
Ecosystem Development: Building localized quantum ecosystems is crucial for fostering innovation and expertise in the field.
The optimism about the potential for quantum computing to revolutionize industries through its integration with HPC is strong, while acknowledging the challenges that lie ahead. The insights shared by the panelists underscore the importance of collaboration and innovation in advancing the capabilities of quantum technologies.
.svg){kind=small}
