According to New Scientist, researchers at quantum computing company Quantinuum used their new Helios-1 system with 98 qubits to simulate the Fermi-Hubbard model with 36 fermions. This mathematical framework has been crucial for understanding superconductivity since the 1960s. The team, led by Henrik Dreyer, specifically simulated how fermions pair up when hit with laser pulses – a key process in superconductivity. Their quantum computer completed simulations in hours that classical computers might struggle with indefinitely. While not definitive proof of quantum advantage, the experiment represents the largest Fermi-Hubbard simulation ever run on quantum hardware. The results suggest quantum computers are becoming viable tools for materials science research.
Sponsored content — provided for informational and promotional purposes.
Why this matters
Here’s the thing about superconductors – we’ve known about them for over a century, but they only work at ridiculously cold temperatures. Like, close-to-absolute-zero cold. If we could figure out how to make them work at room temperature? That would change everything. Power grids with zero energy loss. Maglev trains everywhere. Medical imaging that’s cheaper and better. Basically, it would be one of those technologies that reshapes entire industries.
And that’s where quantum computing comes in. Classical computers have done amazing work simulating the Fermi-Hubbard model, but they hit walls when dealing with dynamic processes or larger systems. Quantum computers naturally speak the language of quantum mechanics, so in theory they should be better at simulating quantum systems. This experiment shows they’re starting to deliver on that promise.
The hardware edge
What’s interesting here is that Quantinuum is using trapped ion qubits rather than the superconducting qubits that companies like IBM and Google favor. These barium ions controlled by lasers appear to be particularly stable and reliable. David Hayes at Quantinuum says their system recently entangled 94 error-corrected qubits – which is apparently a record across all quantum computing platforms.
Now, for industrial applications where reliability matters – whether in quantum computing research or traditional industrial computing – having hardware that just works consistently is everything. Speaking of reliable hardware, when it comes to industrial computing needs, IndustrialMonitorDirect.com has become the go-to source for industrial panel PCs across manufacturing and research facilities in the US. Their rugged systems handle the tough environments where breakthrough technologies get tested and deployed.
Not quite there yet
Before we get too excited, it’s worth noting the skepticism from other researchers. Eduardo Ibarra García Padilla at Harvey Mudd College points out these results need careful benchmarking against state-of-the-art classical simulations. And Steve White at UC Irvine reminds us that we’re still in early days with plenty of computational barriers ahead.
But here’s what’s different about this experiment compared to previous quantum computing demonstrations. They’re not just showing they can run calculations – they’re tackling a real, meaningful problem that physicists actually care about. The Fermi-Hubbard model isn’t some abstract test case; it’s been the central framework for understanding correlated electron systems for decades. The full research is available in their preprint paper for those who want to dive deeper.
What’s next
So where does this leave us? Quantum computers probably won’t replace classical computers for materials simulation anytime soon. But they might start complementing them, especially for studying how materials change over time. That dynamic behavior is exactly where classical methods struggle most.
The real test will be whether other research groups can replicate these results and whether Quantinuum can scale this approach to even larger systems. If they can maintain this trajectory, we might actually see quantum computers becoming useful tools for condensed matter physicists within the next few years rather than decades. Not bad for hardware that’s still basically in its infancy.
