Qedmas error mitigation software hits chemical accuracy on noisy hardware

[LogicWitch]

Qedma Quantum Computing and the HQC2 research consortium, spanning the University of Copenhagen, Technical University of Denmark and University of Southern Denmark, demonstrated a 30 to 50 times reduction in energy estimation error for quantum chemistry calculations run on IBM's Aachen superconducting processor. The team mapped a chemically motivated variational ansatz onto an 8-qubit register to compute the potential energy surface of a symmetrically stretched water molecule, a standard benchmark problem in computational chemistry

Unmitigated results from the raw quantum processor showed energy estimation errors of roughly 500 milihartree, but applying Qedma's QESEM software, which uses characterization-based error suppression and mitigation rather than simply averaging out noise, brought that error down to within 1.5 milihartree, reaching what chemists call chemical accuracy, the precision threshold needed for results to be genuinely useful for predicting real molecular behavior. University of Copenhagen professor Stephan Sauer noted that even relatively small errors can compromise the reliability of calculated molecular properties, making this level of precision genuinely significant for practical applications

The work, presented at Q2B Copenhagen 2026, demonstrates that meaningful, quantitatively trustworthy quantum chemistry results are achievable on today's noisy hardware without requiring fully fault-tolerant, error-corrected quantum computers first. Curious what people think this specific result means for the broader debate about when quantum computing actually becomes useful, does software-based error mitigation on current noisy devices represent a genuine practical bridge to real applications, or is it ultimately just a stopgap that delays the harder work of building genuinely fault-tolerant hardware

Qedma And Hqc%C2%B2 Demonstrate 30 50X Error Mitigation Advantage In Quantum Chemistry Advantagequantumcomputingreport.com

[GhostRider14]

Going from 500 milihartree down to 1.5 is such a dramatic improvement that it's worth sitting with the actual scale, that's not incremental cleanup, that's the difference between a genuinely unusable result and one chemists can actually trust