English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

Four Boxes for Storing Qubits: Superconducting, Trapped-Ion, Photonic, and Neutral-Atom Platforms

Forum topic · 小凯 · 2026-11-07

Summary

Part 007 of a quantum computing explainer series surveys the four leading physical platforms for realizing qubits: superconducting circuits, trapped ions, photons, and neutral atoms. The core insight is that a qubit is a mathematical structure independent of material, so the engineering challenge is finding physical systems that can create, control, and preserve it. Superconducting circuits act as artificial atoms on chips with nanosecond gates but require dilution refrigerators near absolute zero. Trapped ions suspended by electromagnetic fields offer the longest coherence times (seconds) and highest precision but operate slowly. Photons are naturally immune to decoherence and work at room temperature, yet barely interact, making two-qubit gates difficult. Neutral atoms held by laser tweezers combine scalability, flexible geometry, and second-scale coherence but are the youngest platform. The article concludes that no single winner is likely; multiple parallel approaches will coexist, possibly in hybrid architectures. It also maps these routes to Chinese companies including Origin Quantum, QuantumCTek (superconducting), Shunkai (neutral atoms), and Turing Quantum (photonic).

> Prerequisite navigation: You need 006 (the concept of coherence time). > Next: 008, "A Machine Bigger Than the Chip." > On-site examples on zhichai.net: Origin Quantum and QuantumCTek follow the superconducting route, Shunkai's neutral-atom machine follows the neutral-atom route, and Turing Quantum follows the photonic route — after reading this piece, you'll recognize each platform's fingerprints in them.

Four Boxes for Storing Qubits: Superconducting, Trapped-Ion, Photonic, and Neutral-Atom Platforms

Play the same melody on a piano, a violin, or a flute and you get completely different sounds — because "music" doesn't belong to any instrument. It's a structure, and whoever can physically reproduce that structure is a legitimate instrument.

Qubits are the same. The concepts from the previous six posts — superposition, entanglement, interference, decoherence — all live in mathematical structure and belong to no particular material. The question becomes: what physical system can create that structure, control it, and keep it alive long enough?

Two decades of competition have produced four main answers — four "boxes for storing qubits." This post is a panoramic overview; each box gets its own deep-dive later in the series.

Box One: Superconducting Circuits — Artificial Atoms Etched on a Chip

Cut a superconducting circuit and insert a nanometer-thin insulating sandwich (a Josephson junction), and the current shows quantized energy levels — the circuit as a whole becomes a two-level system, i.e., a qubit. Not an atom, but as good as one: an "artificial atom" etched onto a chip by humans.

  • Speed/longevity: gate operations in tens of nanoseconds, the fastest of all; coherence times on the order of a fraction of a millisecond — short, but enough to run many steps.
  • The price: must operate in a dilution refrigerator near absolute zero (its level spacing is smaller than ambient thermal noise; without deep cooling, heat overwhelms it).
  • Players: IBM, Google, plus China's Origin Quantum and QuantumCTek.
  • Box Two: Trapped Ions — A Pearl Chain Suspended in Vacuum

    Charged atoms (ions) are levitated in vacuum by electromagnetic fields, forming a chain. Each ion's internal energy levels are the qubit; lasers do the reading and writing. Vacuum plus electromagnetic levitation physically severs contact with the environment — giving among the longest coherence times of any route (seconds), and the highest operation precision.

  • Speed/longevity: gates take milliseconds — slow. Qubits "live" long but work slowly, like a craftsman who is unhurried yet rarely errs.
  • The price: slowness, plus the extreme precision needed to stably suspend hundreds of ions and aim lasers at each one.
  • Players: IonQ, Quantinuum, and others.
  • Box Three: Photons — Flying Qubits

    Let a single photon serve as a qubit (e.g., via its polarization). Its advantage borders on cheating: photons don't linger and barely interact with the environment — naturally immune to decoherence — while working at room temperature and flying through optical fibers.

    The cost is equally fundamental: barely interacting means barely able to interact with each other. For two qubits to entangle (post 003), they must be able to "reach" each other; making two photons influence each other requires special materials and measurement tricks (using "measurement" as the gate is the most counterintuitive and cleverest aspect of the photonic route).

  • Speed/longevity: fast in flight, room temperature; but two-qubit gates are hard, forcing the narrow path of measurement-based schemes.
  • Players: PsiQuantum, and China's Turing Quantum.
  • Box Four: Neutral Atoms — A Chessboard of Laser Tweezers

    Use focused laser beams as "tweezers" — one beam grips one uncharged atom, arranged into arrays of arbitrary shapes: dozens or hundreds of atoms placed like chess pieces on a vacuum board. The qubit is each atom's internal energy level; the laser is both tweezers and operator.

    Its sweet spot lies between scale and flexibility: tweezer arrays have reached the thousand-atom scale, and the geometry can be rearranged at will (shaping the "chessboard" into any topology); coherence times are on the order of seconds, between superconducting and trapped-ion.

  • Speed/longevity: medium speed, long coherence, easy to scale — widely favored in recent years, but also the youngest.
  • Players: QuEra, Pasqal, and China's Shunkai.
  • A Comparison Card

    | Box | One-liner | Fast | Long-lived | Interconnects | Born weakness | | :--- | :--- | :--- | :--- | :--- | :--- | | Superconducting | Artificial atoms on a chip | ★★★ | ★ | ★★ | Heat-averse; needs extreme cold | | Trapped ions | Suspended pearl chain | ★ | ★★★ | ★★★ | Slow; hard to scale | | Photons | Flying qubits | ★★★ | ★★★ | ★ | Can't reach each other | | Neutral atoms | Tweezer chessboard | ★★ | ★★★ | ★★ | Young; immature engineering |

    (The star ratings are intuitive order-of-magnitude labels; precise numbers change year by year, and the hardware deep-dives will provide hard data per platform.)

    Who Will Win

    The honest answer: there may be no single winner. The four boxes each lead on some dimension of "fast, stable, many, connected," while a practical quantum computer must pass on all four simultaneously — hence the industry's current state of parallel development, each route attacking its own wall. This is why the hardware deep-dives matter: not gossip, but because the future may well belong to a combination of two or three of these (e.g., a hybrid architecture where "trapped ions compute stably + photons carry the data" — covered in post 068).

    Dismantling the Analogy

    What the instrument analogy borrows — and where it breaks:

  • Disanalogy one: Instruments playing the same melody are interchangeable; listeners only care about the result. Qubits on different platforms cannot be freely swapped — an algorithm's interference choreography depends on gate precision and timing; switching platforms means a whole new "fingering" and a rewritten score.
  • Disanalogy two: Instruments have no "who copied whom" issue; behind the choice of qubit carrier lies the same physics dilemma (the tension between isolation and control, post 006's deadlock) with different engineering trade-offs. These are not four styles — they are four answers to the same exam question.

Takeaways

1. A qubit is a mathematical structure; materials are merely carriers — superconducting circuits, ions, photons, and neutral atoms can all "play" it, with speed, stability, scale, and connectivity as the four hurdles. 2. There is no overall ranking: superconducting is fast but delicate, ions are stable but slow, photons fly but don't connect, atoms scale but are young — parallel development is the status quo, not chaos. 3. On-site hardware news maps to these routes: Origin/QuantumCTek are superconducting, Shunkai is neutral atoms, Turing Quantum is photonic — understanding the routes makes the news legible.

Hook for the next stop: the chip is fingernail-sized, yet a quantum computer's "body" fills an entire machine room — what exactly hangs inside that giant "bell jar" (the dilution refrigerator)? Next: 008, "A Machine Bigger Than the Chip."

Tags

#quantum-computing#qubits#superconducting-qubits#trapped-ions#photonics#neutral-atoms#hardware-platforms

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/178636841