Researchers at Duke Quantum Center published results on July 27, 2026, confirming a decades-old theoretical prediction about how entanglement behaves near a quantum critical point, and did it on a real quantum processor rather than in simulation. The team, Thomas Barthel, Marko Cetina, Qiang Miao, Tianyi Wang, and Kenneth R. Brown, ran the experiment on a fully-connected trapped-ion computer and tracked how entanglement entropy scales as a system passes through a phase transition.
What "log-law scaling" means
Physicists have long predicted that at a quantum critical point, the entanglement entropy of a subsystem grows logarithmically with the size of that subsystem, not linearly and not saturating at some fixed value. That distinction matters because it is one of the signatures physicists use to identify and classify phase transitions in quantum systems. Confirming that the log-law holds on an actual device, rather than in a tensor-network simulation of what a device should do, is a different kind of evidence. A simulation assumes the physics it is testing. A quantum computer running the real dynamics does not.
Why full connectivity mattered here
The experiment ran on a trapped-ion system with all-to-all qubit connectivity, meaning every qubit interacts directly with every other qubit rather than only its physical neighbors. Most current quantum hardware, including most superconducting chips, only supports nearest-neighbor connections, which forces long-range correlations to be built up through chains of local operations. That adds noise and depth to exactly the kind of circuit this experiment needed. All-to-all connectivity let the team probe long-range entanglement structure directly, without laundering it through extra gates.
The tomography trick that made small qubit counts work
Studying a genuine phase transition normally requires a system large enough that boundary effects do not distort the result, which is a hard requirement to meet on any current quantum processor. The Duke team worked around this by combining the multiscale entanglement renormalization ansatz (MERA), a tensor-network method built to represent systems as if they were infinite, with holographic subsystem tomography to reconstruct the entanglement structure from a comparatively small number of qubits. That combination is arguably as much the result here as the log-law confirmation itself: a method for getting infinite-system physics out of a finite, noisy device.
What this is not
This is not a demonstration of quantum advantage, and the paper does not claim one. Log-law scaling of entanglement at criticality is a well-established theoretical result, tested here to validate that a real device reproduces it, not to solve a problem no classical computer handles. The value is methodological: a technique (MERA plus holographic tomography) with a path toward studying quantum phase transitions and critical phenomena that are harder to simulate classically, in systems too large for exact classical treatment.
What to watch next
The real test of this method is whether it gets applied to a phase transition or many-body system where the classical answer is not already known, which is where a validated experimental technique starts producing new physics rather than confirming old theory. Our hardware overview tracks how trapped-ion connectivity compares with superconducting and neutral-atom approaches for problems like this one.