Quantinuum and IonQ are both building trapped-ion quantum computers — trapping individual charged atoms and controlling them with lasers or microwaves instead of the superconducting circuits IBM and Google use — but where the last two entries in this series showed a race with no single winner, the IonQ-versus-Quantinuum comparison splits even more cleanly: IonQ is winning decisively on commercial traction and raw revenue, while Quantinuum is ahead on the harder technical bottleneck of turning noisy physical qubits into reliable logical ones. Both companies would tell you they're winning. Based on the numbers, they're each right about a different part of the race.
Why Trapped Ions Are a Different Bet Than Superconducting
Before comparing IonQ and Quantinuum directly, it's worth revisiting why trapped-ion technology exists as a separate category from the superconducting approach IBM and Google use, covered in the first entry of this series. Trapped-ion qubits offer all-to-all connectivity — any qubit can interact directly with any other — and coherence times measured in seconds rather than the microseconds typical of superconducting circuits. That translates into a real efficiency advantage: a trapped-ion system with 1,000 physical qubits could theoretically yield around 500 logical (error-corrected) qubits, compared to roughly 200 for neutral-atom systems and fewer than 10 for superconducting surface-code systems at the same physical scale.
The tradeoff is speed. Trapped-ion two-qubit gates run at roughly 1 millisecond, compared to about 40 nanoseconds for superconducting gates — a gap of more than four orders of magnitude that limits how many operations a trapped-ion machine can run per second, even with its efficiency advantage per qubit.
IonQ: Highest Fidelity, Fastest Revenue Growth
IonQ's core technical claim is the highest raw gate fidelity ever published across any quantum modality: 99.99% two-qubit gate fidelity, announced in October 2025 using a technique called evaporative qubit cooling (EQC) — meaning roughly one error in every ten thousand operations. IonQ runs a dual-species architecture using both barium and ytterbium ions, and its commercial systems (Forte, Forte Enterprise, and Aria) are available through Amazon Braket, Microsoft Azure Quantum, and Google Cloud, giving it the broadest cloud distribution of any trapped-ion vendor. The company's roadmap targets a 256-qubit demonstration system within 2026.
IonQ has also been aggressive on acquisitions: a $1.075 billion purchase of Oxford Ionics in 2025 — the largest trapped-ion M&A deal to date — brought microwave-gate technology in-house alongside IonQ's existing laser-gate approach, and a pending $1.8 billion acquisition of SkyWater Technology, announced in January 2026 and expected to close by mid-2026 pending regulatory approval, would give IonQ its own semiconductor-scale fabrication capacity for trap chips. On the business side, the results back up the aggression: IonQ reported roughly $130 million in full-year 2025 revenue, and Q1 2026 revenue alone hit $64.7 million, a 755% year-over-year jump — more than twelve times Quantinuum's disclosed Q1 2026 revenue of $5.2 million.
Quantinuum: Best Logical Qubit Efficiency, Still Private
Quantinuum's story is almost the inverse — less commercial noise, stronger claims on the hardest unsolved technical problem in the field. Its newest system, Helios, converts 98 physical barium-ion qubits into 48 logical qubits using a technique called color-code encoding, a roughly 2:1 physical-to-logical ratio that the company describes as the best encoding efficiency in the industry. Quantinuum's QCCD (quantum charge-coupled device) architecture physically shuttles ions between processing zones rather than relying purely on optical addressing, a design built for wafer-scale manufacturing of multi-zone trap chips down the line. Backing that up, Quantinuum's H2 processor achieved a quantum volume of 2²⁵ — 33,554,432 — in September 2025, the highest quantum volume ever recorded on any platform, reflecting a combination of gate fidelity and full qubit connectivity that's difficult to fake with marketing.
Quantinuum remains majority-owned by Honeywell and doesn't publicly disclose full revenue figures, but it raised $600 million in September 2025 at a $10 billion valuation, and filed confidentially for an IPO in January 2026, reportedly targeting a valuation around $20 billion. That IPO push, combined with the Helios and H2 results, positions Quantinuum as the company betting its story on being right about hardware before it needs to be right about revenue.
Head-to-Head Comparison
| Metric | IonQ | Quantinuum |
|---|---|---|
| Two-qubit gate fidelity | 99.99% (EQC, Oct 2025) — industry record | 99.9%+ consistently across all qubit pairs |
| Flagship system | Forte Enterprise / Aria (AQ 35+ benchmark) | Helios: 98 physical → 48 logical qubits (2:1 ratio) |
| Notable record | Highest published two-qubit fidelity of any modality | Highest quantum volume ever recorded (2²⁵ on H2) |
| 2026 roadmap target | 256-qubit demonstration system | Continued logical-qubit scaling via QCCD architecture |
| Cloud availability | Amazon Braket, Azure Quantum, Google Cloud | Own Nexus platform with Guppy programming language |
| Q1 2026 revenue | $64.7 million (+755% YoY) | $5.2 million (undisclosed full figures) |
| Ownership / public status | Publicly traded (NYSE: IONQ) | Majority-owned by Honeywell; confidential IPO filing, Jan 2026 |
| Recent major acquisition | Oxford Ionics ($1.075B, 2025); SkyWater pending ($1.8B) | None disclosed at this scale |
So Which One Is "Winning"?
It depends entirely on what you're measuring, and pretending otherwise misrepresents both companies. On commercial traction, cloud reach, and revenue growth, IonQ isn't just ahead — it's more than twelve times larger on quarterly revenue, with real cloud partnerships and government-facing contracts to show for it. On the harder technical bottleneck of the entire quantum industry — converting noisy physical qubits into reliable logical ones without needing an impractical number of physical qubits per logical qubit — Quantinuum's Helios and H2 results currently represent the strongest public evidence in trapped-ion computing, and arguably in the industry overall alongside Google's error-correction work covered earlier in this series.
For a business or research team deciding where to run quantum workloads today, IonQ's broader cloud availability makes it the more accessible starting point. For anyone tracking which company is closer to solving fault tolerance itself, Quantinuum's encoding efficiency numbers deserve equal attention, IPO hype aside. As with the broader quantum race, there's no single scoreboard here — just two different bets on what matters first.
Frequently Asked Questions
Which company has higher gate fidelity, IonQ or Quantinuum?
IonQ holds the published record at 99.99% two-qubit gate fidelity, announced in October 2025 using evaporative qubit cooling. Quantinuum's H-Series achieves 99.9%+ two-qubit fidelity consistently across all qubit pairs, which is extremely close in practical terms despite IonQ's headline number.
Is Quantinuum publicly traded like IonQ?
No. IonQ is publicly traded on the NYSE under the ticker IONQ. Quantinuum remains majority-owned by Honeywell and filed confidentially for an IPO in January 2026, reportedly targeting a valuation around $20 billion, but it had not completed a public listing as of mid-2026.
What does "logical qubit" mean and why does the ratio matter?
A logical qubit is a group of physical qubits combined through error-correction encoding to behave as one reliable, low-error qubit. A better physical-to-logical ratio means fewer physical qubits are needed to get useful, error-corrected computation, which is why Quantinuum's roughly 2:1 ratio on Helios is considered a significant technical achievement.
Why are trapped-ion systems slower than superconducting ones like Google's or IBM's?
Trapped-ion two-qubit gates take about 1 millisecond to execute, compared to roughly 40 nanoseconds for superconducting gates — a difference of more than four orders of magnitude. Trapped ions make up for this with higher fidelity and better connectivity per qubit, but raw operation speed remains their biggest disadvantage.