QCBRIEF
A Journal of the American Chemical Society (JACS) paper shows OTI Lumionics's Iterative Qubit Coupled Cluster algorithm running 200 plus qubit OLED emitter simulations on a single commodity CPU.
A single commercial AMD CPU with roughly 800 GB of RAM just emulated more than 200 qubits of quantum-style chemistry, and the results agreed with lab measurements across 14 phosphorescent compounds used in OLED displays.
The work, published this week in the Journal of the American Chemical Society (JACS) by materials-discovery firm OTI Lumionics and Samsung Advanced Institute of Technology (SAIT), runs OTI's Iterative Qubit Coupled Cluster (iQCC) algorithm, a classical simulation of a quantum-style method, on strongly correlated triplet states in Ir(III) and Pt(II) emitters, the molecules behind efficient red, green, and blue pixels. On an optimized C++ build using 32 processes, the authors reported a mean absolute error of 0.05 electron volts and an R² of 0.94 against experiment. On a NVIDIA Blackwell system, OTI said the same workload ran about 90 times faster than on the CPU box, bringing a 112-qubit ground-state energy calculation down to roughly one hour.
"This represents a paradigm shift," said Dr. Scott Genin, OTI's VP of Materials Discovery, arguing that accuracy is "no longer limited by the size of the hardware." Dr. Tommy Ohyun Kwon called iQCC a "foundational framework" for accelerated materials design.
The benchmark numbers are the authors' own. No independent re-runs of iQCC against the same 14-emitter set have surfaced. The preprint is on arXiv as 2512.13657; the JACS manuscript is the version of record.