QedmaIsrael at the Quantum Frontier
Extending the Frontier of Useful Quantum Computation Through Error Mitigation
Qedma and IBM used QESEM on an IBM Heron processor to study quantum dynamics at up to 74 qubits in a regime where the classical methods tested no longer provided controlled, consistent predictions.
One of the most interesting problems in quantum computing today is that hardware is already capable of running complex computations, but noise can blur exactly the information we are trying to extract from them.
Qedma’s approach is to address part of that problem in the software layer.
The company develops QESEM, an error-mitigation system designed to extract more accurate results from existing quantum processors.
In July 2026, Qedma, IBM, RIKEN, BlueQubit and additional partners published a large-scale study using QESEM on an IBM Heron R3 processor to study a many-body Floquet Ising magnet.
The researchers measured magnetization dynamics with percent-level precision and identified long-lived quantum oscillations in systems of up to 74 qubits.
As the system size and physical time increased, the advanced classical methods tested began to enter a regime where they could no longer provide controlled, consistent predictions: tensor-network simulations did not converge, and other methods remained highly sensitive to approximations even after extensive computations on GPUs and the Fugaku supercomputer.
The quantum computation, after error mitigation, continued to reveal the physical structure of the system.
Just as importantly, the researchers did not rely on a single result. They used several layers of validation, including independent error-mitigation methods, noise-model checks and selected validations on Quantinuum’s trapped-ion hardware.
Why it matters
As quantum computers enter regimes where it becomes difficult to compute the answer classically, a new question becomes central: how will we know that the quantum result is correct?
The work by Qedma and its partners points to an important model for this era — not necessarily validating every answer with a classical computer, but building a chain of checks that makes it possible to trust the computational process and the result.
IBM and Qedma present the result as a demonstration of quantum advantage. The circuits and results were also released for open examination, so the comparison with classical methods can continue to evolve as the field advances.
Technology
The result highlights the importance of an error-mitigation layer between quantum hardware and the application.
Instead of looking only at qubit count or raw processor quality, the practical capability of the system can be improved through software that extracts more accurate information from noisy runs.
In this case, QESEM enabled an existing quantum processor to be used as a computational instrument for studying many-body physics in a regime that is especially challenging for classical computation.
Business significance
If a software layer can expand the type and scale of problems that can be run reliably on existing hardware, it gains a central role in the quantum stack.
The value is not only in improving a processor benchmark, but in turning existing hardware capability into more useful computational capability for researchers and organizations.
Qedma demonstrates something broader about the future of the industry: some of the meaningful advances on the path to quantum value may come not only from a new hardware generation, but also from software layers that make it possible to extract much more from the hardware already available.