Quantum Computing Stocks Split as DARPA Advances IonQ and IBM

The week the quantum race got a referee

The sales pitch for quantum computing stocks hasn't changed much: it's early, the market will be enormous, and the winners will be picked soon. In early October 2026 the people actually doing the picking turned out to be government evaluators, not retail investors. DARPA moved Atom Computing, Diraq, IBM and IonQ into the final phase of its Quantum Benchmarking Initiative (QBI), where government teams will verify and validate the proposed systems hands-on.4 Each of the four uses a different hardware approach: neutral atoms, silicon spin qubits, superconducting circuits and trapped ions. All four came from the 11 companies named to Stage B in November 2025.4 They join Microsoft and PsiQuantum, which reached the equivalent final phase through an earlier DARPA program.4

The stock market didn't respond the way the hype would predict. On the day of the DARPA news, IonQ fell 4.5%, and D-Wave, Rigetti and Quantum Computing Inc. each lost more than 3%. The 10-year Treasury yield briefly hit 5.36%, its highest level since 2002, and that hit speculative tech stocks hard.2 IonQ released its DARPA agreement after the market closed, so the drop wasn't a verdict on the deal.2 Still, the sequence makes the point of this article. The technology is clearly improving, but these stocks still trade mostly on interest rates and sentiment.

What the DARPA selection is worth

IonQ's Stage C agreement could be worth up to $300 million. Any money beyond the initial obligation depends on future appropriations and DARPA funding decisions.10 The evaluation runs through 2029 and will cover several generations of IonQ's Superion systems.10 Analysts at B. Riley described the awards as likely performance-based and possibly tied to appropriations. They kept a Buy rating and a $100 price target.8 B. Riley also pointed out that Rigetti is still in Stage A, while D-Wave and Infleqtion are not disclosed participants.8

This matters more than most "top stocks to buy" lists suggest. The QBI exists to test whether a quantum computer can reach utility scale, meaning its computing value is greater than its cost, by 2033.210 Investors who can't judge qubit physics now have a credible outside filter. Among publicly traded pure plays, IonQ is the only one that has passed it so far. That gap is real, and the bullish sector roundups that rank IonQ, Rigetti and D-Wave as near-equals tend to skip over it.

Error correction, not qubit counts, is the new scoreboard

The technical coverage agrees on one shift. The field has stopped racing for raw physical qubits and now competes on logical qubits, which spread information across many physical qubits so errors can be caught and corrected.1214 The biggest recent result came from IBM and the University of Chicago. They ran encoded circuits on 70 logical qubits, performing 2,415 logical two-qubit operations and 468 logical T gates in about 15 minutes. Effective error rates were ten times lower than the underlying physical rates.20 The researchers said the problem is beyond the practical reach of leading classical simulation methods. They also stressed that the result comes with statistical evidence it can be trusted.20 One University of Chicago researcher called verification one of the biggest obstacles to firmly establishing quantum advantage.20

IonQ's September result was a different kind of advance. The company said it built the industry's first end-to-end real-time error-correction decoder that runs on a single off-the-shelf CPU.19 Decoding can slow a machine down: if the classical processor can't keep up, the quantum hardware has to stop and wait. IonQ says its design avoids that.15 The stock rose more than 11% the next day.15

The details deserve more caution than the trading suggested. IonQ's own release says it tested the decoder on benchmark circuits that simulated up to 408 logical qubits and more than 31.5 million operations, adding as little as 0.02% extra run time.13 That is useful engineering evidence that classical control won't become the limiting factor. It is not a 408-logical-qubit machine. Some market coverage described it as a "408-qubit milestone."3 Even an enthusiastic explainer conceded the advance does not mean fault-tolerant machines exist today.16

Where the reporting diverges

The sharpest disagreement is over how to count logical qubits. One widely circulated guide calls QuEra the record holder with up to 96 verified logical qubits, using a high-rate code published in Nature.11 Another industry ranking credits QuEra with two error-corrected logical qubits. It says the 96 were encoded in blocks where only 0.24% to 3.4% of runs were kept.6 The first guide itself separates codes that only detect errors and throw out bad runs from codes that actually correct them.11 The headline numbers sound alike, but the claims behind them differ. Investors comparing companies on logical-qubit leaderboards should check which kind of claim they're reading.

The coverage also disagrees about corporate facts. One ranking still lists IonQ's $1.8 billion SkyWater acquisition as pending.6 Investor coverage describes IonQ raising its 2026 revenue guidance to $450–460 million including SkyWater's contribution.24 It also reports IonQ fabricating its first integrated 256-qubit processors with SkyWater.7 Our reading is that the deal has moved further than the slower-updating ranking shows, but readers should confirm it in IonQ's own filings.

Commentators even disagree with each other at the same outlet. One Motley Fool contributor called IonQ his top October pick and said it could be a "genius investment" at roughly half its all-time high.26 A week later, another contributor said none of the pure plays had proven enough for him to buy. He cited IonQ's adjusted EBITDA loss of $120.3 million in the second quarter, up from $36.5 million a year earlier.5

The financial reality behind the three usual names

IonQ is clearly the commercial leader among pure plays. Second-quarter revenue rose 287% to $80.1 million, although acquisitions supplied much of that growth.21 It holds about $3 billion in cash and trades at roughly 59 times sales. The tech sector average is about 8.21 Its stated roadmap is aggressive. Commissioning of its 256-qubit system starts in the first half of 2027. It targets 200,000 physical and more than 8,000 logical qubits in functional testing by 2028, then 2 million physical and 80,000 logical qubits by 2029.8 CEO Peter Chapman has pointed to profitability around 2030.21

Rigetti is a smaller company. Second-quarter revenue was $5.1 million, mostly from sales of its 9-qubit Novera systems. Operating expenses rose 48% to $30.3 million, mainly because R&D spending grew.7 An $8.4 million system for India's C-DAC is expected to count as revenue in the fourth quarter, which shows how much single contracts move its results.7 The shares have lost about 31% this year, and Zacks rates the stock Strong Sell.7 Its CEO says the goal is quantum advantage in about three years.21

D-Wave has the most enterprise deployments of the group, using its annealing machines for optimization problems.6 It added gate-model hardware by buying Quantum Circuits Inc. for $550 million.6 Customers include AT&T, Nasdaq's Verafin and NTT DOCOMO. D-Wave and Rigetti each signed agreements for up to $100 million in CHIPS Act funding.24 But D-Wave's second-quarter revenue slipped to about $3.1 million against an adjusted EBITDA loss of $37.1 million, and the stock trades near 485 times sales.21 It is also near the bottom of its 52-week range.3

The incumbents and the toolmakers

Two large companies complicate the idea that investors must pick a pure play. IBM made DARPA's final stage, produced the 70-logical-qubit result, and plans to spend $10 billion over five years on its goal of a large-scale fault-tolerant machine by 2029.520 CEO Arvind Krishna has said quantum could noticeably affect IBM's revenue and earnings by 2028 or 2029.5 The stock trades at about 20 times earnings.21

Nvidia has said it is not building a quantum processor. Instead it sells the connecting pieces: CUDA-Q software, the NVQLink interconnect and AI models for error correction.22 Its reach is already visible. IonQ's Superion 256 will be the first platform deployed at Nvidia's Accelerated Quantum Research Center.24 Infleqtion used an early version of Nvidia's CUDA-Q Logical to validate a code that needs about six physical qubits per logical qubit.18

The verdict

Read together, the coverage leads to a clear conclusion. The technical progress of 2026 is real. Error correction is becoming practical, verification is getting stricter, and the timelines to useful machines are clustering around 2029.12 But the "top three stocks" framing is already out of date. The more useful question is which companies are passing independent tests. Right now that list includes IonQ and IBM, and it does not yet include Rigetti or D-Wave.8 Until revenue catches up with valuations, interest rates will move these stocks at least as much as research results do. October's selloff on the day of a major contract showed that clearly.2