Xanadu uses photons, IonQ uses trapped ions, and Rigetti uses superconducting circuits. Those different qubit technologies shape how each company controls its hardware, connects components, and plans to scale. The evidence available through 2026 does not establish a universal winner: performance figures depend on the particular system, gate, benchmark, and measurement method.
How the three approaches differ
| Company | Computational medium | Engineering approach | What to keep in mind |
|---|---|---|---|
| Xanadu | Photons, or particles of light | Photonic hardware and a networked, modular direction; its Aurora demonstration connected photonic racks by optical fiber. | Claims about scalability, energy efficiency, and computational advantage are Xanadu’s own and should not be treated as independent comparative findings. |
| IonQ | Trapped ions: individual atoms held in place | Lasers prepare and measure the qubits; the system uses vacuum, optical, and control infrastructure. | IonQ presents all-to-all connectivity and high fidelity as advantages of its approach. These are vendor claims, not proof that it outperforms the alternatives for every workload. |
| Rigetti | Superconducting circuits | Superconducting processors, including modular chiplet designs, with cryogenic infrastructure. | Its reported results are specific to named processor generations and metrics; one processor’s figure should not be applied to another. |
These are different ways of building and controlling qubits, not three interchangeable product configurations. The physical medium matters, but practical performance also depends on the full system, the workload, and how a result is measured.
As an Amazon Associate I earn from qualifying purchases.
How each company’s hardware works
Xanadu: computing with light
Xanadu’s 2026 Form F-1 describes photons as the computational medium and presents a full-stack strategy that pairs photonic hardware with PennyLane, an open-source quantum programming framework. The company says PennyLane supports circuits across multiple hardware modalities and cloud platforms.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Scan for outdated or missing drivers - takes under a minute3Repair Windows errors before they cause bigger problemsXanadu identifies Borealis and Aurora as demonstrations. Its filing describes Borealis as a 216-qubit photonic system used for a 2022 computational-advantage demonstration. It also says Aurora demonstrated real-time error detection and optical-fiber interconnection between photonic racks. These descriptions are company-reported.
#1 Best Overall
The filing estimates that simulating the computation Borealis performed in two minutes would have taken the Fugaku supercomputer approximately seven million years. That is Xanadu’s estimate for that specific computation; it does not establish a general speed advantage for useful applications.
Xanadu’s filing also describes long-term physical- and logical-qubit targets and a target architecture dated 2029–2030. Those are roadmap goals, not demonstrated or delivered systems.
IonQ: computing with trapped atoms
IonQ describes using naturally occurring individual atoms as qubits, trapping them in three-dimensional space, and using lasers to prepare and measure them. The company’s technology account describes the associated vacuum, optics, and control infrastructure. IonQ highlights all-to-all connectivity and high fidelity as strengths of its systems; the comparative value of those properties depends on the task and the specific hardware.
Free tools Windows power users keep installed
One-click scans. No signup required.
In September 2026, IonQ announced the Superion product line and Electronic Qubit Control. The announcement included a planned Superion 256 system and expected customer deliveries in 2027. IonQ explicitly characterizes statements about future development and delivery as forward-looking, so these should be read as plans rather than available capabilities.
Rigetti: computing with superconducting circuits
Rigetti describes superconducting quantum processors and modular chiplet designs in its 2026 Form 10-K. Its systems require cryogenic infrastructure. The company also offers Novera, a specialized nine-qubit research processor based on its Ankaa-class architecture. Novera is intended for research and development, and its product materials specify a compatible dilution refrigerator and laboratory setup; it is not a consumer device.
What the published performance figures do—and do not—show
Qubit counts and fidelity percentages are not standalone rankings. A meaningful comparison needs the hardware generation, gate type, calibration conditions, benchmark definition, date, and whether the figure is internal testing or independently validated. The cited company materials do not provide a matched independent comparison across all three vendors.
Rank #4
| System or claim | Reported figure | Context and attribution |
|---|---|---|
| Xanadu Borealis | 216 qubits | Xanadu’s 2026 Form F-1 describes the system used for its 2022 computational-advantage demonstration. |
| IonQ two-qubit gate fidelity | 99.99% | IonQ reports this as a 2025 result, restated in 2026 company materials and tied to a particular technology. The cited material does not establish a matched independent comparison with Xanadu or Rigetti. |
| Rigetti Cepheus-1-36Q | 99.6% median two-qubit gate fidelity; 76-nanosecond median gate time | Rigetti’s 2026 Form 10-K reports these figures from internal testing as of January 2026. |
| Rigetti Cepheus-1-108Q | 108 qubits; 99.1% median two-qubit CZ gate fidelity | Rigetti’s technical page lists deployment on April 7, 2026. This is a different system and measurement context from Cepheus-1-36Q. |
Even two fidelity figures expressed as percentages may refer to different gates, procedures, and system conditions. A higher headline percentage or larger qubit count alone does not show which machine will perform better on a particular task.
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Scan for outdated or missing drivers - takes under a minuteDriver Scan →How to choose a fair comparison
Start with the workload or research question, then compare systems using the same evaluation criteria. Useful questions include:
- What is the physical qubit medium? Photons, trapped ions, and superconducting circuits involve different hardware and control choices.
- How are components connected? IonQ claims all-to-all connectivity; Xanadu describes fiber-linked photonic racks; Rigetti describes chiplet-based processors. These descriptions apply to the companies’ particular systems and designs, not automatically to every machine using that modality.
- What does the performance number measure? Check the gate type, fidelity definition, gate time, test conditions, date, and whether the result was internal or independently validated.
- Is the capability demonstrated, deployed, or planned? A demonstrated research result, a deployed cloud system, specialized lab equipment, and a roadmap target are different levels of availability.
- What has been shown about scaling and error correction? Separate working demonstrations from long-term architecture goals. Current noisy quantum systems are not interchangeable with fault-tolerant systems capable of useful large-scale computation.
How to access the platforms
For software experimentation, Xanadu’s PennyLane is open source and designed to work across hardware modalities. IonQ lists access routes through AWS, Microsoft Azure, Google Cloud, and Nvidia. Rigetti describes its Quantum Cloud Services platform and public-cloud access. Availability, account requirements, and the hardware exposed through a cloud provider can vary, so check the provider’s current offering before choosing a platform.
Rigetti’s Novera is a separate institutional option for groups that need on-premises research equipment. Its cryogenic and laboratory requirements make it a specialized procurement decision rather than a general-purpose way for individuals to try quantum computing.
Is one approach best?
Not on the evidence presented here. Xanadu’s photonic strategy, IonQ’s trapped-ion systems, and Rigetti’s superconducting processors make distinct engineering trade-offs, and their public performance claims are not aligned in a way that supports a single overall ranking. The useful choice depends on the workload, access route, system generation, and the quality of comparable evidence—not on a qubit count or headline fidelity figure in isolation.
Recommended Free Tools
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
