In a feat of computational muscle, researchers at Lawrence Berkeley National Laboratory have used nearly the full power of the Perlmutter supercomputer — 7,168 NVIDIA GPUs running simultaneously — to create the most detailed electromagnetic simulation of a quantum processor chip ever produced.
The achievement, led by Quantum Systems Accelerator researchers Zhi Jackie Yao and Andy Nonaka, marks a significant leap forward in how scientists design and validate quantum hardware before it's ever manufactured.
11 Billion Grid Cells, One Tiny Chip
The chip itself measures just 10 millimeters across and 0.3 millimeters thick, with features as small as one micron. But to capture its electromagnetic behavior in full detail, the team discretized it into 11 billion grid cells — an astonishing level of precision that required nearly every GPU on one of the world's most powerful supercomputers.
"I'm not aware of anybody who's ever done physical modeling of microelectronic circuits at full Perlmutter system scale," said Nonaka. "We were able to run over a million time steps in seven hours, which allowed us to evaluate three circuit configurations within a single day."
Beyond the Black Box
What sets this simulation apart from previous efforts is its refusal to take shortcuts. Many quantum chip simulations treat the device as a "black box," simplifying internal structures because the computational cost of modeling them is too high. The Berkeley team took the opposite approach.
"We care about what material you use on the chip, the layout of the chip, how you wire the metal — the niobium or other type of metal wires — how you build the resonators, what's the size, what's the shape," said Yao. "We include them in our model."
By solving Maxwell's equations in the time domain, the simulation captures how electromagnetic signals propagate through the chip in real time, including nonlinear effects and the subtle interactions between qubits and surrounding circuitry.
Why This Matters for Quantum Computing
Quantum computers hold extraordinary promise for solving problems in drug discovery, materials science, cryptography, and artificial intelligence. But building reliable quantum hardware remains one of the greatest engineering challenges of our time. Unwanted electromagnetic interference — known as "crosstalk" — between qubits is one of the biggest obstacles.
Detailed simulations like this one allow researchers to predict and prevent crosstalk before a chip is fabricated, potentially saving months of trial-and-error manufacturing. The computational model predicts how design decisions affect electromagnetic wave propagation, ensuring proper signal coupling and avoiding interference.
The work was supported by the National Energy Research Scientific Computing Center's Quantum Information Science program and used ARTEMIS, an exascale modeling tool developed under the Department of Energy's Exascale Computing Project. It represents not just a breakthrough in quantum computing research, but a stunning demonstration of what supercomputers can do when pointed at problems that genuinely demand their scale.
