IBM and Israel's Qedma Say They've Pushed a Quantum Computer Past Classical Supercomputers

IBM and Qedma Quantum Computing say they have run a quantum simulation that classical computers can no longer keep up with — a claim that, if it holds, marks one of the more concrete steps toward quantum machines doing useful work.

According to reporting summarized by MSN, IBM (NYSE:IBM) and Qedma announced research that they say demonstrates a quantum computing advantage over leading classical simulation methods, including calculations performed on conventional systems. The Quantum Insider frames the result as an error-mitigated quantum simulation that goes beyond classical benchmarks, while HPCwire reports that Qedma's error-mitigation software is what extends the quantum computer's reach past classical simulation.

That software piece is the heart of the story. Today's quantum processors are noisy: their qubits drift and make mistakes fast enough that long calculations dissolve into garbage. Full error correction — the eventual fix — is still years out. Error mitigation is the pragmatic middle path, using software to characterize the noise and mathematically subtract its effects from the results. Qedma, an Israeli startup, builds exactly that layer, and Globes describes the partnership as a quantum breakthrough for IBM and the Israeli company.

The Qedma work was not announced alone. MSN reports that IBM disclosed key quantum advances on Thursday across three separate collaborations — with the University of Chicago, with Qedma, and with an Italian partner not identified in the available summaries. Citybiz likewise describes three research breakthroughs.

One caution worth keeping: quantum advantage claims have a history of being narrowed or contested once classical researchers sharpen their own algorithms in response. The available reports describe what IBM and Qedma say their results show, and the companies are the source of the comparison.

It matters because the bottleneck in quantum computing has always been noise, and this is a claim that clever software — not just better hardware — can buy real computational ground today.