Nvidia runs 104-qubit quantum simulation at Jülich supercomputer

Nvidia (NASDAQ: NVDA) announced it has completed what it describes as the largest comprehensive simulation of a fault-tolerant quantum computing system to date, running a 104-qubit simulation on JUPITER, Europe's first exascale supercomputer, located at Jülich Supercomputing Centre in Germany.

The simulation ran on 8,192 NVIDIA GH200 GPUs and was used to validate a technique called magic-state cultivation, a fault-tolerant quantum computing approach that the company says requires 90% fewer physical qubits than previous methods. Prior simulations had modeled up to 50 qubits under ideal, noiseless conditions; this simulation modeled quantum processors under realistic hardware noise.

Magic states are specific qubit arrangements that are continuously consumed during fault-tolerant quantum computation. Optimizing how they are produced reduces the resources required to build a functional quantum computer.

"Useful quantum computing will require fantastically complex, fault-tolerant systems — and understanding how they behave at scale is essential to building them," said Jensen Huang, founder and CEO of Nvidia. "Our results from Jülich show how quantum processors and GPU supercomputers can work together to tackle research that would otherwise be out of reach."

Kristel Michielsen, director of the Jülich Supercomputing Centre at Forschungszentrum Jülich, said the collaboration is "helping scientists make better-informed decisions about the designs most likely to deliver useful applications."

Development and testing of the simulation methods were conducted at Nvidia's Accelerated Quantum Research Center using an Nvidia GB200 NVL72 system, scaling from four GPUs to hundreds before moving to JUPITER. The process reduced a weeks-long simulation effort to days.

The simulations used Nvidia's cuQuantum software and the CUDA-Q platform's fault-tolerant logical layer. According to the press release, the results are intended to help researchers refine quantum hardware designs before physical construction, with potential applications in drug discovery, financial modeling, and materials science.

You May Also Be Interested In

• Active options: NVDA CRWV AAPL AMZN TSLA SPCX MU AMD INTC IREN
• Asana stock rises on recent analyst support and market strength
• HPE plans to deploy AI and HPC computing on the Moon in 2026