Researchers at the Jülich Supercomputing Centre and NVIDIA have achieved a major milestone in quantum computing by fully simulating a universal quantum computer with 50 qubits for the first time. The accomplishment was made possible using JUPITER, Europe’s first exascale supercomputer, which was officially launched at Forschungszentrum Jülich last September.
The achievement surpasses the previous record of 48 qubits, which was also set by Jülich scientists in 2019 using Japan’s K computer. Beyond setting a new benchmark, the breakthrough highlights the enormous capabilities of JUPITER and could accelerate the development of future quantum algorithms and technologies.
Why Quantum Simulations Matter
Simulations of quantum computers play a critical role in advancing quantum research. Scientists use them to test algorithms, validate experimental findings, and explore how future quantum systems may behave before real hardware becomes powerful enough to handle such tasks.
Some of the algorithms researchers are interested in include the Variational Quantum Eigensolver (VQE), which can help study molecules and materials, and the Quantum Approximate Optimisation Algorithm (QAOA), designed for solving optimization problems in areas such as logistics, finance, and artificial intelligence.
The Enormous Challenge of Simulating Quantum Systems
Recreating a quantum computer on a traditional supercomputer is extremely demanding because the complexity grows exponentially with every added qubit. Each new qubit doubles both the memory and computing power required for the simulation.
A standard laptop can manage simulations involving roughly 30 qubits. Simulating 50 qubits, however, requires around 2 petabytes of memory, which equals roughly two million gigabytes.
“Only the world’s largest supercomputers currently offer that much,” says Prof. Kristel Michielsen, Director at the Jülich Supercomputing Centre. “This use case illustrates how closely progress in high-performance computing and quantum research are intertwined today.”
The simulation models the detailed quantum behavior of an actual processor. Every operation — such as applying a quantum gate — influences more than 2 quadrillion complex numerical values, a “2” with 15 zeros. Those values must remain synchronized across thousands of computing nodes to accurately reproduce the behavior of a real quantum processor.
NVIDIA GH200 Superchips Enabled the Record
The breakthrough relied heavily on NVIDIA GH200 Superchips used within the JUPITER system. These chips tightly connect central processing units (CPUs) and graphics processing units (GPUs), allowing data that exceeds GPU memory capacity to be temporarily stored in CPU memory while maintaining high performance.
To take advantage of this architecture, engineers at the NVIDIA Application Lab — a initiative between the Jülich Supercomputing Centre (JSC) and NVIDIA — upgraded Jülich’s quantum simulation software, the Jülich Universal Quantum Computer Simulator (JUQCS). The updated version, called JUQCS-50, can efficiently perform quantum calculations even when some of the data is transferred to CPU memory.
Researchers also introduced a byte-encoding compression technique that cuts memory requirements by a factor of eight, along with a dynamic optimization system that continually improves data exchange between more than 16,000 GH200 Superchips.
“With JUQCS-50, we can emulate universal quantum computers with high fidelity and tackle questions that no existing quantum processor can yet solve,” says Prof. Hans De Raedt of the Jülich Supercomputing Centre and lead author of the study published as a preprint.
Expanding Access to Quantum Research
JUQCS-50 will also be made available to outside research organizations and companies through JUNIQ — the Jülich UNified Infrastructure for Quantum Computing. Researchers expect it to serve both as a scientific tool and as a benchmark for evaluating future supercomputers.
The project was developed as part of the JUPITER Research and Early Access Programme (JUREAP). “Through early collaboration, hardware and software could be co-designed during JUPITER’s construction phase, in close cooperation between Jülich experts and NVIDIA — an important step towards realising the full potential of this exascale system,” explains Dr. Andreas Herten, a member of the Jülich JUPITER project team and co-author of the study.
JUPITER receives joint funding from multiple organizations. Half of the funding comes from the European High Performance Computing Joint Undertaking (EuroHPC JU). One quarter is provided by the Federal Ministry of Research, Technology and Space (BMFTR, formerly BMBF), while the remaining quarter comes from the Ministry of Culture and Science of the State of North Rhine-Westphalia (MKW NRW) through the Gauss Centre for Supercomputing (GCS).