Nvidia Expands Quantum Computing with Global CUDA-Q Deployments and HPC Advances
NVIDIA is making strides in advancing high-performance computing (HPC) and quantum computing by leveraging its technology in multiple supercomputing sites and quantum computing centers across the globe. The chipmaker's Grace Hopper Superchips are now being utilized in nine new supercomputers worldwide, contributing to an aggregated processing capability of 200 exaflops, equivalent to 200 quintillion calculations per second. These superchips are lauded for their energy efficiency, with claims of twice the efficiency compared to traditional x86 systems with GPU support.
In addition to these advancements in HPC, NVIDIA is also focusing on quantum computing by promoting the use of its open-source CUDA-Q platform. This platform is being adopted by supercomputing sites in Germany, Japan, and Poland to facilitate the integration of quantum processing units (QPUs) into their systems. QPUs operate on the principles of quantum mechanics, using particles like electrons or photons to perform calculations in a fundamentally different way from traditional processors, potentially enabling faster and more complex computation.
One of the key installations is at Germany's Jülich Supercomputing Centre, where a QPU built by IQM Quantum Computers will complement the existing JUPITER supercomputer. Similarly, the National Institute of Advanced Industrial Science and Technology in Japan is enhancing its ABCI-Q supercomputer with a QUP from QuEra to advance its quantum computing initiative. Poland's Poznan Supercomputing and Networking Center has installed two photonic QPUs built by ORCA Computing, connected to a new supercomputer partition powered by NVIDIA Hopper.
These developments signify a substantial leap forward in the fields of scientific research and discovery, with the potential to tackle complex problems in areas such as weather and climate modeling, materials science, and other domains that require immense computational resources.
