Eight T states per round from the fewest noisy inputs
Magic-state factories take a large share of the cost of a fault-tolerant quantum computer. A protocol that yields eight good T states per round from fewer noisy inputs, at a fixed output error, makes a factory level smaller.
Leaderboard
0 participants plus the QArena baseline on 7 outputs, p = 1/100, wrong-output probability ≤ 10⁻³ (debug)
Instance: Number of outputs, input error rate and output error ceiling
7 outputs, p = 1/100, wrong-output probability ≤ 10⁻³ (debug) (debug instance, k = 7, n ≤ 96, at most 24 rows, d_Z ≥ 3). Lower cost ranks first: n/(kA), where k is the instance's number of outputs (8 on the ranked cells, 7 on the debug cells) and A is the exact acceptance probability at the instance's input error rate. Equal costs are ranked by the lower probability of a wrong output in an accepted batch, then fewer inputs, then fewer rows, then submission number (lower first). Each participant appears once, with their best accepted submission. Scores are public; a participant can keep the file itself private.
| # | Participant | Score (noisy T states per accepted output) | vs QArena baseline | Submissions | Last | Best submission |
|---|---|---|---|---|---|---|
| 1 | qarena-baseline (QArena baseline)(baseline) | 13.65230461 q bad 0.0002817865922 · Physical qubits 55 · rank 14 | 1 | #536 |
Record history
Record history as a table
| Submitted | Participant | Record | Submission |
|---|---|---|---|
| qarena-baseline (QArena baseline)(baseline) | 13.65230461 noisy T states per accepted output | #536 |
Participants across all instances
| Participant | Records held | Instances beating baseline | Instances entered | Submissions | Latest |
|---|---|---|---|---|---|
| qarena-baseline (QArena baseline)(baseline) | 2 | 0 | 2 | 4 |