Zhejiang University 125-qubit surface-code processor
Zhejiang University + Graduate School of China Academy of Engineering Physics (Beijing)
Zhejiang University 125-qubit surface-code processor
Register profile — the machine's variant of its architecture
qubit type: transmon
gate mechanism: tunable nearest-neighbour coupling, CZ
connectivity: planar 2-D array
control: room-temperature RF
control placement: room-temp
readout: dispersive; data-qubit idle error 0.024 per measurement cycle
gate evidence: a published figure (source class not yet tagged)
decoder mode: offline — syndromes decoded after the run
codes: rotated_surface
flags: post-selection dominates the result
Technologies by layer — the machine's cell per layer
| 1 Qubit carrier | |
| 2 Encoding | Bare physical qubit (no encoding) · primary ◌ 🔎 bare transmon |0⟩/|1⟩ subspace; family-determined; leakage out of it post-selected away in the QEC runs |
| 3 Gate mechanism | Tunable-coupler CZ / iSWAP · primary ▣ Fig. 1c (CZ Pauli error CDF); §II ✅ tunable nearest-neighbour coupling, CZ; median CZ fidelity 0.996, median 1Q fidelity 0.9995 (Jun 2026) |
| 4 Connectivity / transport | Static nearest-neighbour lattice · primary ▣ Fig. 1a (connectivity), Fig. 1b (patches on square grid) ✅ planar 2-D array of 125 transmons with tunable nearest-neighbour coupling (degree 4) |
| 5 Control | Room-temperature electronics + per-qubit coax/flex · primary ▤ 🔎 room-temperature RF control; cycle time not stated |
| 6 Readout | Dispersive microwave readout (+TWPA, Purcell) · primary ▣ Fig. 1c (two-/three-state readout error CDFs); §II ✅ dispersive; readout error 0.025 (two-state) / 0.049 (three-state discrimination); data-qubit idle error 0.024 per measurement cycle |
| 7 Code | Rotated surface code (+ yoked variants) · primary ▣ Fig. 1b,d (d=3 patches, merge/split, memory); §II ✅ rotated surface: two d=3 patches with lattice merge/split; logical error/cycle 0.0365 / 0.0282 after leakage rejection (82.4 % of runs discarded); Bell ≥0.532 Magic-state factory (cultivation / distillation / code switching) · alternate ▣ Fig. 4a (logical magic-state preparation); §V 🔎 magic-state injection + gate teleportation on rotated-surface patches: logical R_X(π/4) at 0.943; d=7 teleportation transition (Jun 2026) |
| 8 Decoder | MWPM / Sparse Blossom (+ correlated matching) · primary ▤ §III (belief-matching decoding of Bell-state data) 🔎 belief-matching decoder run offline; MWPM is the nearest Atlas node (no belief-matching node) — none — nothing in this layer ▤ §III (belief-matching decoding of Bell-state data) 🔎 belief-matching, offline: no Atlas node for belief-matching and no real-time decoding |
| 9 Interconnect | |
| 10 Manufacturing | Superconducting-qubit lithography (Al/AlOx/Al junctions, Nb wiring, 300 mm) · primary ◌ 🔎 superconducting-qubit lithography; family-determined, no machine-specific fab source |
Standard records
acceptance fraction → 6.00e-4 · worst published (0.06% retention on the balanced Deutsch-Jozsa instance)
17.6% of runs survive leakage rejection · Jun 2026 · source
2Q error under full-width parallel operation → 4.00e-3 · typical (99.6% median CZ)
Jun 2026 · source
idle / memory error per round → 0.024 · typical (data-qubit idle error per cycle)
Jun 2026 · source
SPAM error → 0.025 · typical (two-state; 4.9e-2 three-state)
Jun 2026 · source
evidence: ✅ 5 / 12