Back to feed
News Story
DeepTech深科技
1 sources

Princeton team achieves millisecond-scale atom reloading while preserving qubit coherence

A research team led by Jeff Thompson at Princeton University published a study in Science demonstrating fast, continuous, and coherent atom reloading for neutral-atom quantum processors. Using mobile optical tweezers to extract atoms from a reservoir, they achieved a characteristic loading time of about 1 millisecond—nearly two orders of magnitude faster than the previous 90-millisecond coherent reloading experiment—without observable additional decoherence on existing qubits. This advance could overcome atom-loss limits on quantum circuit depth and enable sustained operation of neutral-atom quantum computers.

SynthePulse Insight · AI deep readingMembers

Millisecond-Scale Atomic Resupply: How Neutral-Atom Qubits Can Be Swapped "On the Fly"

Version 1 · 1 source

Jeff Thompson's team at Princeton University reported in Science a "coherent reloading" scheme: replenishing new atoms from a reservoir into optical tweezers within about 1 millisecond, while no additional effect on the lifetime and coherence of already-stored qubits was observed. This provides the conditions for long-circuit experiments no longer limited by the initial number of atoms, but there is still a clear distance from full fault-tolerant quantum computing.

  • The paper was published in Science, titled "Fast, continuous, and coherent atom reloading in a neutral-atom qubit array," with Li Yiyi, Yicheng Bao, and Michael Peper as co-first authors, and Jeff Thompson as corresponding author.
  • The characteristic time to load atoms from the reservoir into optical tweezers is about 1 millisecond, and transport also takes only about 1 millisecond; a previous two-species reloading experiment that preserved coherence took about 90 milliseconds.
  • In continuous extraction tests, a 256-site optical tweezer array performed "load, transport, release" at up to 500 times per second, corresponding to extracting 73,000 atoms per second.
Open section navigationThe Problem: Atoms Get Lost, and Replenishing Them Causes Disturbance

The Problem: Atoms Get Lost, and Replenishing Them Causes Disturbance

Neutral-atom quantum processors have made progress in scale and operation precision in recent years, and researchers have demonstrated quantum error correction and algorithms based on logical qubits. But if one can only use a batch of atoms prepared before computation begins, then as atom loss and leakage errors accumulate, the depth of executable quantum circuits becomes limited. Background gas collisions, atom transport, gate operation errors, and some readout methods can all cause atom loss.

Replenishing atoms itself is not easy: it often requires magnetic field gradients and cooling, and imaging light, which may disrupt the states of existing qubits. A previous two-species reloading experiment that preserved quantum coherence had a characteristic loading time of about 90 milliseconds. How to simultaneously achieve reloading speed and coherence is precisely the problem this research aims to solve.

The Scheme: A Reservoir Plus Moving Optical Tweezers, Making "Atom Swapping" a Continuous Process

The team set up a continuously replenished high-density cold-atom reservoir near the computation region, used moving optical tweezers to extract atoms from it, and then sent them to subsequent operation regions. Optical tweezers are tiny optical traps formed by focused lasers that can capture and move atoms; an array composed of multiple optical tweezers can process many sites simultaneously.

This speed depends on a continuously operating atom supply system: cold atoms start from a three-dimensional magneto-optical trap in an adjacent vacuum chamber, are guided about 35 centimeters along a cavity-enhanced optical dipole trap to the reservoir, and are then decelerated and cooled by one-dimensional optical molasses. The time constant for establishing the reservoir is about 5 seconds, but once steady state is reached it can supply continuously, without needing to refill from scratch after each extraction.

The optical tweezer array used for extraction is generated by crossed acousto-optic deflectors and operates at a wavelength of 488 nanometers. During loading, the array overlaps with the reservoir; after loading, a galvanometer mirror moves it to an imaging region about 300 micrometers away to reduce interference from atomic fluorescence in the reservoir. During imaging, the optical molasses is briefly turned off; atoms in the reservoir can still be retained, but their average replenishment rate is affected by the duty cycle of the molasses being on.

Free for now

Read the full analysis

4 more sections of analysis, plus the full takeaway

Loading

Credibility boundary

This article is based on a WeChat public account report (DeepTech深科技, source_tier 2) and the Science paper information it relays. The paper publication, author attribution, loading time, extraction rate, coherence comparison values, and other details all come from that report's relay of the paper and are marked as confirmed in the evidence pack; however, this article did not independently consult the original Science paper, and readers who need to cite specific values are advised to check the original paper. Regarding future cycle rates of hundreds of times per second, reduced fault-tolerance overhead, and benefits for atomic clocks and quantum simulations, the source only states "may" or "goals yet to be achieved," and this article treats them as source_claim, not as already realized.

Primary report

DeepTech深科技

Primary source