IonQ, Rigetti, D-Wave, Quantinuum: Same Sector, Five Bets, Zero Winners

That's the headline everyone wrote. The real story underneath it is messier: the market is pricing quantum computing as if there's one winning architecture and one clean commercialization path. There isn't. Five genuinely different physical approaches are racing at once, none has decisively pulled ahead, and the most aggressive timeline claim in the group is also the one drawing the sharpest pushback from the physicists who study this for a living.

Start with the architectures, because the ticker soup makes no sense without them. $IONQ and the newly public $QNT (Quantinuum, majority-backed by $HON) both trap individual charged atoms with electromagnetic fields and manipulate them with lasers. It's the slowest gate speed of the group, roughly 1,000 to 10,000 times slower than superconducting gates by Rigetti's own estimate, but it holds the cleanest error profile and the highest quantum volume result on record: Quantinuum's H-Series hit 33,554,432 in September 2025. $IONQ is the only pure-play name with real, audited commercial revenue growth to show for it. Q1 2026 revenue came in at $64.7 million, up 755% year over year, and management raised full-year guidance to $260 to $270 million with remaining performance obligations up 554% to a record $470 million. $QNT's Q1 2026 revenue, by contrast, fell 73% year over year to $5.24 million as the company reorganizes ahead of going public, though full-year 2025 revenue had grown to $30.9 million from $23 million the year before. Both companies are burning heavily: IonQ's adjusted EBITDA loss guidance for 2026 sits between $310 million and $330 million, and Quantinuum posted a $136.5 million net loss in the first quarter of 2026 alone.

$GOOGL, $IBM and $RGTI all bet on superconducting qubits: tiny loops of metal cooled near absolute zero, fast gates, shorter coherence times than trapped ions. This camp produced the only claim on record that other researchers have said is genuinely different from marketing hype. $GOOGL's 105-qubit Willow chip ran an algorithm called Quantum Echoes, an out-of-time-order correlator measurement across a 65-qubit subsystem, about 13,000 times faster than the best classical estimate on a supercomputer, and published the result in Nature in October 2025. Google Quantum AI's Hartmut Neven and Vadim Smelyanskiy called it "the first time in history that any quantum computer has successfully run a verifiable algorithm that surpasses the ability of supercomputers." That's a meaningfully different claim than Google's 2019 "quantum supremacy" run, which used a problem custom-built to be hard for classical computers and useless for much else; this one was applied, in a proof-of-concept with UC Berkeley, to actually measure molecular structure via NMR.

$RGTI's approach is modular: twelve linked 9-qubit chiplets forming its 108-qubit Cepheus-1-108Q system, now generally available across Rigetti's own cloud service, Amazon Braket, Microsoft Azure Quantum and qBraid. It currently runs at 99.1% median two-qubit gate fidelity at roughly 60-nanosecond gate speeds, with a company roadmap targeting 99.5% fidelity later this year. Commercially, Rigetti is still tiny: Q1 2026 revenue was reported around $4.4 million against more than $569 million of cash on hand, alongside a system-sale agreement with India's Centre for Development of Advanced Computing.

$MSFT went in a completely different direction: topological qubits built from a theoretical quasiparticle called a Majorana zero mode, which in principle protects information from errors at the hardware level rather than leaning on heavy error-correction overhead. In June, Microsoft unveiled Majorana 2, reporting topological qubit lifetimes exceeding 20 seconds, more than 1,000 times longer than the 1-to-12-millisecond lifetimes of its first-generation Majorana 1 chip, achieved largely by swapping aluminum for lead in the superconducting stack. Microsoft used the result to pull its fault-tolerant computer timeline forward from 2033 to 2029. Physicists were not convinced: demonstrating a real qubit requires two types of measurement, called X and Z, and the Majorana 2 paper presents only the Z measurements, the same gap critics flagged after Majorana 1 eighteen months earlier. Henry Legg of the University of St Andrews put it bluntly: "Nothing in this preprint resolves the fundamental issues." Microsoft's own quantum lead, Chetan Nayak, is standing behind the 2029 date regardless, saying of the coherence jump: "We're seeing this more than 1000x improvement in this critical metric of the qubit."

Then there's $QBTS on quantum annealing, a fundamentally different computing model built for optimization rather than general-purpose gate operations, and $QUBT on photonic qubits encoded in light. Both look odd next to the gate-based names because their headline numbers this year came mostly from deal-making rather than steady operations. $QBTS's Q1 2026 revenue fell 81% year over year to $2.9 million, but only because the year-ago quarter included a one-time $12.6 million system sale; bookings actually rose to a record $33.4 million, up nearly 2,000% year over year, anchored by a $20 million Florida Atlantic University purchase and a $10 million Fortune 100 services deal, and the company closed the acquisition of Quantum Circuits Inc. to build out its own gate-model offering. $QUBT's revenue jumped to $3.7 million from about $39,000 a year earlier, a headline gain north of 9,000%, but nearly all of it came from consolidating two acquisitions, Luminar Semiconductor and NuCrypt; stripped of those deals, organic revenue was roughly $24,000.

None of these five bets is close to being proven wrong or right, and the valuations don't reflect that uncertainty so much as ignore it. A foreignpolicyjournal analysis of the group in early July put $IONQ's price-to-sales multiple near 109x, $RGTI's near 836x, and $QBTS's near 791x; $IONQ carries a beta near 3.23, meaning it's built to move roughly three times the market's daily swing in either direction, and it has. By late July, after another leg down, those multiples had compressed hard but stayed extreme: IONQ near 65 times sales, and D-Wave, Rigetti and Quantum Computing Inc. still between roughly 385 and 480 times revenue. The whole group ripped through April and May on strong earnings, then gave much of it back. IonQ fell 21% in a single week in early June despite posting record revenue that same period; the same analysis clocked June drawdowns of 25% for IonQ, 24% for Rigetti and 21% for D-Wave on the month, with no company-specific news behind any of it. Another leg down hit mid-July: IonQ fell to about $39 on July 13, down 31% over the trailing month, while Rigetti, D-Wave and Quantum Computing Inc. each slid about 6% the same session on broad risk-off flows rather than any one headline.

$QNT's debut is the cleanest test of what a real institutional-grade IPO looks like landing on top of that volatility. It priced at $60 a share, above an already-raised $53-to-$55 range, selling 28 million shares to raise $1.68 billion. Shares opened at $68, hit a session high of $71.35, then closed essentially flat, valuing the company near $15.7 billion. Honeywell will retain about 48.1% of Quantinuum's combined voting power; founder Ilyas Khan, now chief product officer, holds roughly 15% personally, a stake reportedly worth more than $2 billion at the IPO valuation. Quantinuum CEO Rajeeb Hazra told CNBC that Honeywell would remain "a very disciplined shareholder" and would eventually "deal with this asset, monetize this asset." Nvidia's venture arm, alongside Quanta Computer, JPMorgan Chase, Mitsui, Amgen and others, had already backed Quantinuum in a $600 million private round at a $10 billion valuation the previous September, less than nine months before the public price came in roughly 50% higher.

You can check pieces of this yourself instead of taking a stranger's word for it. Compare IonQ, Rigetti, D-Wave and Quantum Computing Inc. on revenue growth and margins pulls the four US-listed names side by side. Rank the same group by price-to-sales shows how stretched the multiples actually are — IonQ near 65 times sales, the other three several hundred times revenue. And asking whether any quantum company actually turns a profit surfaces exactly one name, IonQ, and even that only on a GAAP technicality driven by non-operating gains rather than the core business — which tells you most of what you need to know about where this sector sits on the hype curve.

Quantum computing in July 2026 is running on two different clocks. The physics clock moves in years and is genuinely contested: Legg and other critics haven't backed down, and won't until Microsoft publishes the X-measurement data everyone has been asking for since Majorana 1. The market clock moves in days, and it has already rewarded and punished the same handful of stocks for the same unresolved questions multiple times this year. The government taking equity, not just handing out grants, is a real signal that Washington now treats this as closer to a strategic asset than a lab curiosity. It isn't a signal that any of the five architectures has actually won. The next real test isn't the next stock pop. It's whether Q2 bookings across this group hold up once the sentiment swings stop doing the heavy lifting.