Will a Chinese team publish below-threshold surface-code error correction at code distance 9 or more before 31 December 2027?
deadline 2027-12-31
confidence medium
tier 4 evidence
35%Asturna estimate
Resolution criterion
Yes if a peer-reviewed paper or an arXiv preprint from a Chinese institution reports a surface code of distance at least 9 with a measured error suppression factor above 1.
Verified starting state
In December 2025, the USTC team published a cover result in Physical Review Letters on Zuchongzhi 3.2 (107 qubits): below-threshold correction on a distance-7 surface code, suppression factor 1.4. Going from d=7 to d=9 requires roughly 161 data qubits plus ancillas. Google took about two years to go from d=5 to d=7.
What pushes it up
- USTC has demonstrated rapid development cycles, achieving a d=7 below-threshold surface code on their 107-qubit Zuchongzhi 3.2 processor in late 2025 [1].
- The resolution criterion includes arXiv preprints, which significantly reduces the time required to meet the deadline compared to waiting for peer review.
What pushes it down
- Scaling from d=7 to d=9 requires roughly 161 data qubits plus ancillas, a massive jump from the 107 total qubits on Zuchongzhi 3.2 [1].
- Google required approximately two years to transition from d=5 to d=7 [1], and the transition from d=7 to d=9 is expected to be even more complex.
- With only 14 months remaining until the deadline and no preprint yet released, the timeline for fabrication, calibration, and publication is extremely tight.
What would move this number most
Whether USTC or another Chinese group has already fabricated and is currently calibrating a >161-qubit processor capable of maintaining sub-threshold error rates.
How this number is built
| Deadline | 2027-12-31 |
| Base rate | 30% |
| Best evidence | tier 4 · 0 article(s) used |
| Independent runs | 35 · 35 (spread 0 pts) |
| Confidence | medium |
| Evidence cap | not triggered |
| Last revised | 2026-10-08 09:19 |
| Model | gemini-3.5-flash |
Probability history
2026-10-08 09:19 35% 0 article(s) on file · gemini-3.5-flash
- What pushes it up. USTC has demonstrated rapid development cycles, achieving a d=7 below-threshold surface code on their 107-qubit Zuchongzhi 3.2 processor in late 2025 [1].
- What pushes it up. The resolution criterion includes arXiv preprints, which significantly reduces the time required to meet the deadline compared to waiting for peer review.
- What pushes it down. Scaling from d=7 to d=9 requires roughly 161 data qubits plus ancillas, a massive jump from the 107 total qubits on Zuchongzhi 3.2 [1].
- What pushes it down. Google required approximately two years to transition from d=5 to d=7 [1], and the transition from d=7 to d=9 is expected to be even more complex.
- What pushes it down. With only 14 months remaining until the deadline and no preprint yet released, the timeline for fabrication, calibration, and publication is extremely tight.
- What would move this number most. Whether USTC or another Chinese group has already fabricated and is currently calibrating a >161-qubit processor capable of maintaining sub-threshold error rates.
2026-09-22 00:38 35% 0 article(s) on file · openai/gpt-oss-120b (Groq)
Reasons not recorded for this revision (detailed history has only been kept since recently).