ETH Zurich achieves 99.91% quantum gate fidelity across 17,000 atom pairs
Researchers at ETH Zurich have demonstrated a quantum swap gate with 99.91% fidelity operating simultaneously on over 17,000 pairs of neutral atoms. Published in Nature, the work introduces a gate inherently resistant to common experimental flaws, potentially reshaping paths to large-scale, fault-tolerant quantum computing.
The team, led by Professor Tilman Esslinger and principal investigator Konrad Viebahn, built a two-qubit gate using geometric phases rather than finely tuned dynamic phases typical in most quantum platforms. They trapped potassium-40 fermionic atoms in an optical lattice and created transient "qubit doublon" states where two atoms share a site. Fermionic exchange antisymmetry in these doublons yields a two-particle quantum holonomy free of dynamic phases. This protects the swap operation against laser potential fluctuations and inhomogeneities. Additional resilience stems from time-reversal and chiral symmetries in the underlying Hamiltonian.
This result stands out for its scale. While ion-trap systems like IonQ's 99.99% two-qubit fidelity in October 2025 measured single pairs, ETH Zurich verified 99.91% across the entire 17,000-pair system. This highlights neutral atom platforms' scalability advantages. Combined with the group's prior topological pumping for atom transport, it lays groundwork for highly connected, large-scale quantum processors.
Neutral atom quantum computing gains traction alongside superconducting circuits and trapped ions. These findings address a core challenge: making quantum operations robust enough to scale without impractical environmental controls. Esslinger noted the ability to generate vast numbers of swap gates with neutral atoms. The authors call it a new paradigm turning fundamental symmetries and quantum statistics into resources for fault-tolerant computing.
-
16:22
-
16:18
-
15:44
-
15:27
-
15:10
-
14:53
-
14:05
-
14:00
-
14:00
-
13:45
-
13:42
-
13:40
-
13:29
-
13:27
-
13:25
-
13:10
-
12:54
-
12:54
-
12:50
-
12:47
-
12:30
-
12:12
-
11:50
-
11:41
-
11:32
-
11:28
-
11:15
-
11:08
-
11:00
-
10:46
-
10:41
-
10:21
-
10:20
-
10:19
-
10:05
-
09:47
-
09:45
-
09:36
-
09:34
-
09:25
-
09:09
-
08:47
-
08:32
-
08:15