Tsinghua's FaultScope simulator maps where quantum error correction fails

[BeckyLynch]

Finding which parts of an error correction circuit cause the most problems could help engineers improve quantum computers faster. Quantum Zeitgeist reports on FaultScope, a simulator developed by Tsinghua University and the Beijing Institute of Mathematical Sciences and Applications

FaultScope is written in Rust and Python. Along with logical error rates, it calculates how sensitive those rates are to each noisy component in a circuit, producing a hotspot map of where errors matter most

Calculating these sensitivities adds only 0.33 to 0.97 percent to the running time. A distance 100 circuit ran in 2,043 seconds, while the SymFT tool timed out at distance 30 and above. At distance 70, FaultScope took 619 seconds compared with 1,742 seconds for Stim, a widely used simulator

The work is a preprint, and its fixed budget of 100,000 shots may not be enough to resolve very low error rates accurately

Knowing which gates or qubits contribute most to logical errors is valuable, because engineers can focus improvements where they help most. Beating Stim on speed at large distances is a notable claim, since Stim is the field's standard tool. Independent testing will show how well it holds up

How useful are hotspot maps for improving quantum hardware? And can new simulators challenge Stim?

FaultScope Software Pinpoints Errors In Quantum Error Correction Circuitsquantumzeitgeist.com