IBM and Qedma achieve quantum advantage on 74 qubits, beating supercomputers

IBM and Qedma have demonstrated quantum advantage using commercially available hardware, resolving physics problems that defeated one of the world's most powerful supercomputers.

Quantum computers running Qedma's error mitigation software have solved a physics problem beyond the reach of Fugaku, Japan's flagship supercomputer, marking the first time quantum advantage has been achieved with commercially available hardware.

"IBM quantum computers have reached a maturity where they can produce solutions that, for the first time, achieve both trust in the solution through extensive testing and outperform the best classical simulation methods," Jay Gambetta, Director of IBM Research and IBM Fellow, said.

The team used IBM's Quantum Heron processor — available on the cloud — paired with Qedma's Quantum Error Suppression and Error Mitigation (QESEM) software to model the dynamics of a two-dimensional Floquet Ising model across systems of up to 74 qubits. When RIKEN and BlueQubit ran the same problem on Fugaku and other leading classical simulators, none of the classical approaches could consistently agree at the scale the quantum experiments reached. The error-mitigated quantum results showed clear, long-time oscillatory behavior that the classical methods missed.

The breakthrough strengthens IBM's quantum roadmap and Qedma's error correction approach at a time when the quantum computing market is projected to reach $65 billion by 2030, according to McKinsey. IBM, which also recently acquired HRL Laboratories' quantum dot technology, is positioning across multiple qubit modalities — superconducting, spin-based, and now error-mitigated systems — as competitors including Google, Microsoft, and Quantinuum race toward practical quantum applications.

Validation Against Classical Limits

The study's design addressed a long-standing criticism of quantum advantage claims: trust. The team first validated their error-mitigation protocol against classical calculations where comparisons remained possible, then pushed into regimes where classical simulations lost accuracy. They benchmarked a more scalable mitigation approach against those trusted results before extending to larger system sizes and longer evolution times. Independent validation on Quantinuum's trapped-ion hardware produced consistent behavior, suggesting the measured physics came from the simulated quantum system rather than device-specific errors.

Qedma's QESEM software, available in the Qiskit Functions Catalog on the IBM Quantum Platform, uses a patented approach that includes an unbiased error-mitigation method providing guaranteed accuracy. The company was founded by Dr. Asif Sinay together with Prof. Dorit Aharonov, who three decades ago proved that error-corrected quantum computation is possible, and Prof. Netanel Lindner, a condensed matter physicist.

What This Means for the Quantum Race

The demonstration shifts the quantum computing narrative from theoretical promise to practical capability. Rather than waiting for fully fault-tolerant quantum computers — still years away — QESEM enables larger, more complex workloads on today's noisy hardware. The team publicly released their quantum circuits and results on the Quantum Advantage Tracker, inviting continued benchmarking from the research community.

For investors, the implications span multiple tickers. IBM (IBM) strengthens its position in the quantum computing race against Alphabet's Google (GOOGL), which claimed quantum supremacy in 2019 with Sycamore, and Microsoft (MSFT), which has bet on topological qubits. Quantinuum, the trapped-ion leader that validated IBM's results, remains privately held. The ability to run useful computations on 74-qubit systems today, rather than waiting for thousand-qubit fault-tolerant machines, could accelerate the timeline for commercial quantum applications in materials science, drug discovery, and cryptography.

This article is for informational purposes only and does not constitute investment advice.