TL;DR
Recent repetition-code experiments show step changes in protected-bit lifetime as researchers identify and fix failure modes, producing a rapid, superexponential improvement pattern the author calls “FOOM.” The same dynamics are expected to apply to full quantum (logical) qubits as surface-code implementations scale and known hurdles are addressed.
What happened
A sequence of experiments tracking repetition-code protection on superconducting devices recorded dramatic gains in lifetime: a 9-qubit demonstration in 2014 showed a ~100 microsecond half life, a 21-qubit run in 2021 reached ~3 milliseconds, a 51-qubit test in 2023 achieved ~300 milliseconds, and a 59-qubit experiment in 2024 reported a roughly two-hour half life. The author explains these jumps with a compact model where lifetime grows as L = C * λ^q (citing an equation in the surface-code literature), and argues that exponential improvements in qubit count stacked on exponential error suppression produce a lull followed by a rapid “FOOM.” Real-world ceilings—so-called QEC hurdles—interrupt the idealized superexponential growth; identified examples include leakage from transmons and high-energy events like cosmic rays. Targeted fixes such as gap engineering removed a cosmic-ray-related limit in 2024 and produced a large measured lifetime jump, illustrating how removing hurdles leads to further gains.
Why it matters
- If the stacked-exponential dynamics hold, modest yearly increases in qubit counts combined with error-correction scaling can yield very large gains in logical qubit lifetime.
- Identifying and mitigating practical failure modes (QEC hurdles) appears critical to converting theoretical error suppression into measured, usable improvements.
- The same scaling logic that improved repetition-code protected classical bits should, in principle, extend to protecting quantum states with codes like the surface code.
- Success in these areas would change the landscape of what error-corrected quantum hardware can practically achieve, affecting roadmaps and engineering priorities.
Key facts
- 2014: UCSB/John Martinis group used a 9-qubit repetition code to protect a classical bit with ~100 µs half life.
- 2021: A 21-qubit repetition code experiment extended the half life to ~3 ms.
- 2023: A 51-qubit repetition code produced a ~300 ms half life, but measurements deviated from simple scaling due to high-energy impacts.
- 2024: A 59-qubit repetition code achieved about a two-hour half life after mitigations (gap engineering) addressing high-energy events.
- The author uses a model L = C · λ^q (cited to surface-code literature) where q is physical qubit count and λ measures qubit quality.
- In a surface code, quadrupling physical qubits squares the logical error rate; the author cites a 2024 best surface-code logical error of 0.1% per round (~300 µs half life).
- Examples in the source illustrate how successive quadruplings (holding quality constant) would map from 0.1% to 0.0001% (~300 ms), to 1e-10% (~3 days), and further to impractically tiny rates (~30 billion years) absent real-world ceilings.
- Known QEC hurdles include leakage (transmons leaving the qubit subspace) and high-energy particle impacts; fixing such hurdles produces stepwise lifetime improvements.
What to watch next
- Progress on leakage-detection and removal techniques for superconducting qubits.
- Further mitigations for high-energy events (cosmic rays) and other system-level failure modes.
- Scaling experiments that increase physical-qubit counts without degrading per-qubit quality, and reported surface-code logical error-rate trends.
Quick glossary
- Repetition code: A simple error-correction scheme that encodes one logical bit by repeating a physical bit across multiple qubits to protect against bit-flip errors.
- Surface code: A two-dimensional quantum error-correcting code that protects logical qubits by arranging many physical qubits and stabilizer measurements to correct both bit- and phase-flip errors.
- Logical qubit: An encoded qubit formed from many physical qubits under an error-correcting code, intended to be more robust than any single physical qubit.
- Leakage: When a physical qubit exits the computational subspace (e.g., higher energy levels in a transmon), producing errors not corrected by standard qubit-error models.
- Logical error rate: The probability that an error-corrected (logical) qubit experiences an uncorrected error per correction cycle or unit time.
Reader FAQ
What does “FOOM” refer to in this context?
The author uses FOOM to describe a rapid, superexponential improvement in protected-qubit lifetime that follows a period of slow progress when stacked exponentials align.
Have logical (error-corrected) qubits begun to outperform physical qubits?
According to the source, error-corrected objects recently began to surpass individual physical qubits in lifetime measurements.
Is the quality barrier guaranteed to fall within a specific time frame?
The author predicts the qubit-quality barrier may fall within five years, but this is presented as the author’s expectation rather than an established fact.
Will these improvements immediately enable full-scale fault-tolerant quantum computers?
Not confirmed in the source.
Quantum Error Correction goes FOOM 24 Dec 2025 In this post: why I expect the maximum achievable qubit quality to increase drastically in the next few years. In 2014, the…
Sources
- Quantum Error Correction Goes FOOM
- Why Qubits need quantum-error correction (and vice versa)
- New qubit circuit enables quantum operations with higher …
- Quantum Error Correction: Time to Make It Work
Related posts
- Why Calling ‘The Global Market’ a Strategy Is an Irresponsible Shortcut
- A Christmas knock that changed lives: a Cardiff couple’s 45-year guest
- Self-referencing Page Tables for x86 Architecture: Bachelor Thesis Findings