[ISC 2024] NVIDIA's Grace-Hopper offerings adopted by multiple national supercomputing centers together with CUDA-Q platform

Aside from conventional computing adopting NVIDIA Grace-Hopper Superchips for even better supercomputing performance, the quantum computing sector also started to push them into action armed with Team Green's CUDA-Q platform.

Specifically, institutions based in Germany, Japan, and Poland will be using all the hardware and software from NVIDIA to power the Quantum Processing Unit (QPUs) for significant advancements in data processing.

In the case of Germany’s Jülich Supercomputing Centre (JSC), the GH200-supercharged system is receiving a QPU made by IQM Quantum Computers in the form of a complement module for the JUPITER Supercomputer which will be used for things like chemical simulations and optimization-related workflows made possible by the superconducting qubits.

On the other hand, the ABCI-Q supercomputer sitting inside the National Institute of Advanced Industrial Science and Technology (AIST) in Japan is slated to implement QuEra's QPU unit and they look forward to accelerate development in AI-based quantum applications as well as energy and biology breakthrough by utilizing Rubidium atoms controlled by laser light as qubits to perform calculations.

Lastly, Poland’s Poznan Supercomputing and Networking Center (PSNC) took in a pair of photonic QPUs by ORCA Computing, connected to a new supercomputer partition accelerated by NVIDIA Hopper. With 2x PT-1 quantum photonics systems, they are crafted a bit different as they use telecom frequencies as qubits so that the architecture can be scaled and be fully modular just with off-the-shelf telecom equipment. Sectors that the researchers are going to cover with the supercomputer are in biology, chemistry, and machine learning.

Aside from tackling selected field of studies, NVIDIA also believes that its CUDA-Q platform will be able to utilize AI to solve the issues of quantum computing as a tech itself such as noisy qubits as well as efficient algorithms to drive over better output.