IBM Claims Three Quantum Advantage Breakthroughs
Get the Health newsletter
Daily health & science — research, biotech, public health, the studies worth knowing. Free.
- IBM researchers and colleagues claimed three new cases of quantum advantage, each accompanied by explicit methods for verifying how much of the quantum output can be trusted versus how much is noise.
- Fugaku (Japan's flagship supercomputer) and an Nvidia GPU both failed to produce results for a simulation of how a quantum magnet responds to laser light, while the quantum computer completed the calculation and revealed previously unknown oscillating properties in the magnet.
- In a second simulation of how information scrambles within uneven materials — relevant to catalysis chemistry — the quantum computer outperformed leading classical simulation methods at certain parameters.
- Abhinav Kandala at IBM led a trust-quantification approach that ran simulations on multiple quantum computers with varying noise levels and injected controlled noise, separating genuine signal from noise.
- The third example addressed a long-standing verification gap for quantum sampling problems by tweaking and repeating computations to measure the fidelity of answers previously lost to errors.
- Independent reviewers split on implications: Emanuele Dalla Torre at Bar-llan University called the magnet finding "a basic science problem" with no clear real-world use yet, while ETH Zurich's Dominik Hangleiter called the verification work "a step in the right direction."
Why it matters: For the first time in these head-to-head comparisons, classical machines including Fugaku and Nvidia's GPU could not match the quantum result, giving researchers a tested method to trust quantum outputs. The practical effect: scientists studying quantum materials and catalysis can now treat quantum machines as working instruments rather than lab curiosities — though the demonstrated applications remain basic science rather than commercial use.
