Connects decision-makers and solutions creators to what's next in quantum computing
Nvidia Supplies Quantum Software for Japan’s Hybrid Supercomputer
The ABCI-Q system aims to provide researchers and companies with the means to explore quantum computing applications
Nvidia is partnering with Japan's National Institute of Advanced Industrial Science and Technology (AIST) research facility to build a hybrid classical-quantum supercomputer.
AIST is developing the cloud-based service, ABCI-Q, as part of Japan’s national quantum computing initiative and expects the service to be available by 2025.
It will be used by researchers to advance quantum circuit simulation and machine learning, build classical-quantum hybrid systems and develop new algorithms inspired by quantum technology.
Companies will also be able to explore potential applications such as optimizing logistics operations and developing new molecules for novel materials, drug development and the chemicals industry.
By offering access to both researchers and the private sector, AIST intends to advance quantum computing technology from both the systems development and potential applications side.
Classical computers currently outperform quantum computers but this is likely to change for certain use cases within a few years. Even then, quantum computers will continue to tackle practical problems working in combination with classical computers in a hybrid setup to make the most of the
Nvidia announced at its GTC conference in March that it would be providing the GPUs for the supercomputer program and the latest announcement confirms it will also provide quantum computing software via a cloud service.
Nvidia’s CUDA-Q open-source hybrid quantum computing platform offers simulation tools and capabilities to program hybrid quantum-classical systems. The supercomputer part of the ABCI-Q project uses Nvidia H100 Tensor Core GPUs in nodes that are interconnected using the Nvidia Quantum-2 InfiniBand networking platform.
Read more about:
AsiaAbout the Author
You May Also Like