Research

Qunova says HI-VQE reached chemical accuracy on 4Fe-4S cluster using 80 physical qubits

Qunova Computing, based in Daejeon, South Korea, announced on Nov. 12, 2026 that its HI-VQE algorithm reached chemical accuracy at 80 qubits (the release does not say whether physical or logical; the run used an IBM Quantum System Two at RIKEN, so these are most likely physical qubits) while simulating the 4Fe-4S iron-sulfur cubane cluster. The workflow paired the IBM machine with RIKEN's Fugaku supercomputer. Qunova reports a ground-state energy extrapolation of -327.2482 ± 0.0007 Hartree, within chemical accuracy of a DMRG reference value of -327.2471 Hartree, and says it is more accurate than the latest extrapolated TrimCI results. Qunova also says HI-VQE reached a lower energy (-326.9821 Ha) than a previously published sample-based quantum diagonalization result by its second iteration, using 2,518 Fugaku node-hours, a roughly 300x reduction in computational cost. The claims come from the company's press release; the release does not mention peer review.

Why it mattersIron-sulfur clusters like 4Fe-4S are core components of enzymes involved in processes such as nitrogen fixation, and are a hard test case for chemistry methods. Qunova says the result puts a quantum-classical workflow on par in accuracy with leading classical methods for this molecule, but it also says it is still working to find systems where HI-VQE consistently outperforms them.

Read the full story at HPCwire →

Summarized by AI from the linked article.