TheMachine Press

A daily newspaper for the age of artificial intelligence.

Morning editionPermanent story

research

Hardware Noise, Not Expressibility, Stopped the Quantum Simulation

At four lattice sites, 1% depolarizing noise drove mean error as high as 52.81%.

Published Updated Story ID: mp-2026-09-28-016
Read the complete editionStory JSON

Summary

At four lattice sites, 1% depolarizing noise drove mean error as high as 52.81%.

A two-flavor Schwinger-model study scales variational circuits from two to six staggered lattice sites and separates expressibility, trainability and hardware noise. Its noiseless four-site control reached 0.12% mean error, while the same circuit under 1% depolarizing noise reached 25.69% to 52.81%. The result is a model-specific diagnosis, not a bound for all quantum simulations.

Why it matters

At four lattice sites, 1% depolarizing noise drove mean error as high as 52.81%.

Limits and context

  • The result is a model-specific diagnosis, not a bound for all quantum simulations.

Key claims

  1. At four lattice sites, 1% depolarizing noise drove mean error as high as 52.81%.

    Qualification: The result is a model-specific diagnosis, not a bound for all quantum simulations.

    Evidence: source-2026-09-28-018

Sources

  1. arXiv preprint 2609.30496arXiv · primary research

Corrections

No corrections have been recorded for this story.