QMill has achieved record-level quantum circuit compression across both Ion Trap and Rigetti native gate sets, extending a prior breakthrough announced on October 15, 2025. Utilising an AI-powered methodology, the company reduced the GF2^8_MULT benchmark circuit—native to Rigetti processors—from 4,213 gates to 1,597 in one hour, outperforming QUESO’s 24-hour result of 1,739 gates. Similarly, for Ion Trap systems, the qcla_mod_7 circuit was compressed from 2,494 to 1,099 gates, exceeding GUOQ’s 1,324 gate reduction. These advancements address limitations of Noisy Intermediate-Scale Quantum (NISQ) systems by minimising gate count and circuit depth while preserving full circuit functionality.
Expanding Quantum Circuit Compression to New Hardware
QMill has expanded its AI-powered quantum circuit compression technology beyond IBM’s gate sets to now include native architectures from IonQ (Ion Trap) and Rigetti (superconducting). This advancement addresses a critical bottleneck in near-term quantum computing: the limitations of NISQ systems due to noise and coherence. By optimising circuits for specific hardware, QMill aims to enable more complex computations within current hardware constraints, bringing practical quantum applications closer to reality.
Recent benchmarks demonstrate significant compression improvements. For the GF2^8_MULT circuit on Rigetti hardware, QMill reduced the gate count from 4,213 to just 1,597 in one hour – outperforming QUESO’s 1,739 gates achieved after 24 hours. Similarly, on Ion Trap systems, the qcla_mod_7 circuit saw a 56% reduction (to 1,099 gates), exceeding GUOQ’s 47% reduction. These results highlight not only gate minimisation but also substantial runtime improvements.
Crucially, QMill’s compression method maintains full circuit functionality – preserving the unitary transformation – while minimising both gate count and circuit depth. This ensures computational accuracy isn’t sacrificed for efficiency. By continuously adding support for diverse hardware and gate sets, QMill positions itself as a key enabler for maximising the potential of today’s and near-future quantum computers and for developing quantum-advantage algorithms for NISQ computing.
These advancements are significant for Noisy Intermediate-Scale Quantum (NISQ) systems. Limited coherence times and susceptibility to noise necessitate minimising gate counts and circuit depth. QMill’s compression techniques directly address these challenges, bringing the promise of practical quantum computations closer to reality by maximising the potential of current and near-future hardware. Continued expansion to support more hardware providers is planned.
