TUM Unveils RushHour, Running 86% of Quantum Benchmarks on Smaller Chips
Updated
Updated · Quantum Zeitgeist · Aug 22
TUM Unveils RushHour, Running 86% of Quantum Benchmarks on Smaller Chips
1 articles · Updated · Quantum Zeitgeist · Aug 22
Summary
86% of tested quantum benchmarks ran with RushHour on constrained chips where existing methods failed entirely, marking a step toward practical fault-tolerant quantum computing.
RushHour replaces rigid pre-allocation of qubits and routing space with dynamic lattice surgery, reallocating ancilla workspace during computation through a hardware-software co-design.
1.2 to 3.5 times larger processors were previously needed for the same workloads, while RushHour also delivered median speedups of 2.0 to 7.2 versus leading alternative compilers.
Six state-of-the-art compilers were used for comparison, and the system still ran 4.8 times slower than an idealized machine, leaving room for further optimization.
The researchers said future work will test how well the approach holds up under broader chip-quality variations and noise conditions.