IonQ to present nine peer‑reviewed papers at IEEE Quantum Week 2026; four win Best Paper Awards

IonQ will present nine peer‑reviewed papers at IEEE Quantum Week 2026 (QCE26) in Toronto. The research focuses on applying trapped‑ion quantum computers to machine learning, scientific computing, logistics, and computational chemistry. Four of the nine papers received Best Paper Awards.

Overview of presentations

The nine presentations cover protein folding, quantum machine learning, finite‑element simulation, cargo selection, advection‑diffusion equations, clinical data imputation, fine‑tuning foundation AI models, distributed quantum optimization, and error mitigation in computational chemistry. They include experiments on IonQ Tempo, Forte, and Forte Enterprise, as well as quantum–classical hybrid workflows. IonQ researchers will participate in a keynote and in panels, workshops, and tutorials on topics including software–hardware integration for quantum error correction, distributed quantum architectures, and quantum software. The announcement lists presentation titles across seven events, but the body of the release also refers to “six additional events,” indicating a discrepancy in the count.

Key points

In collaboration with Kipu Quantum, IonQ extended protein‑folding optimization on IonQ Tempo to problems of up to 61 qubits, reaching the classical reference energy for 4 of the 6 sequences tested.
In a quantum‑accelerated graph partitioning study with Synopsys, applied to industrial models of up to 35 million elements, end‑to‑end finite‑element simulation wall time was reduced by up to 14.6%.
The logistics study with Einride addressed problems up to 130 qubits and increased the number of selectable shipments by up to 12.1% without a significant cost increase.
With QuantumBasel and the University of Basel, quantum fine‑tuning of a foundation AI model reduced classification error by up to 24% compared with the best classical baseline, and an energy‑to‑solution breakeven point was observed at around 34 qubits.
Using mid‑circuit measurement on IonQ Tempo, researchers addressed error mitigation for computational chemistry problems. In a study of quantum parity representations, a configuration where inference was performed entirely classically still improved accuracy by up to 41.7 percentage points.

Technical and business implications

These presentations are significant because they evaluate trapped‑ion quantum computers not only by raw computational performance but as components within end‑to‑end workflows that combine classical and quantum processing. Specifically, finite‑element work measures end‑to‑end wall time, AI work measures accuracy and energy‑to‑solution, and the logistics studies measure cost and the number of selectable shipments. The collaborations with multiple companies and research institutions, together with four Best Paper Awards, demonstrate results presented under external peer review. However, the announcements do not allow assessment of whether the reported advantages over classical methods will persist in commercial deployments or translate into customer adoption and revenue.

What to watch next

Going forward, it will be important to see how each paper specifies the classical baselines used, total processing time including classical computation, and the evaluation conditions including I/O and pre/post‑processing. Additional indicators will be whether similar improvements hold on different datasets or at larger scales, whether independent third parties can reproduce and benchmark the results, and whether collaborative research progresses from demonstrations to customer deployments or concrete product offerings.

NETWORKView IonQ’s industry network →

Source

Read the original announcement