Research Finder
Find by Keyword
IonQ: Record Q2 and Buying the Foundry Whose Largest Customer Was, Itself
Record $80.1 million quarter, SkyWater closes, RPOs climb 297%, and a $120 million intercompany line that disappears on SkyWater consolidation
08/12/2026
By the Numbers
- Revenue: $80.1 million, up 287% YoY from $20.7 million, roughly 20% above the midpoint of prior guidance and above the $66.4 million Zacks consensus
- Organic revenue growth: 132% YoY per management on the call, against a full-year organic target of 100%
- Adjusted EPS: ($0.33), better than the ($0.54) consensus
- Adjusted EBITDA: ($120.3) million versus ($36.5) million a year ago; ($95.6) million excluding SkyWater-related spend
- GAAP net loss attributable to IonQ: ($1,867.7) million, GAAP EPS ($5.08), driven primarily by roughly $1.6 billion of non-cash warrant mark-to-market
- GAAP operating expenses: $417.3 million, or $201.2 million non-GAAP; R&D was $160.6 million of GAAP OpEx
- Remaining performance obligations: $485 million, up from $470 million in Q1 and $122 million a year ago, growth of 297% YoY
- Revenue mix: approximately 50% international, 60% commercial, 25% multi-product; multi-product sales grew 40% YoY
- Cash, equivalents and investments: $3.0 billion at June 30, 2026; approximately $2.0 billion pro forma for SkyWater
- FY2026 guidance raised to $280 million to $290 million from $260 million to $270 million, IonQ standalone, excluding any SkyWater contribution
- Implied second half of roughly $135 million to $145 million (analyst arithmetic: full-year guide less first-half revenue of $144.7 million, a range raised by an analyst on the call and not disputed by management)
- SkyWater acquisition: $1.8 billion, closed July 31, 2026, at $15.00 cash plus 0.4883 IonQ shares per SkyWater share
- IonQ spending with SkyWater under the commercial agreement: $24.7 million in Q2, with full-year spend forecast at approximately $120 million
The News
IonQ reported its fifth consecutive record quarter with revenues printing at $80.1 million. The company also raised full-year 2026 guidance to a range of $280-290 million. The print landed well above consensus on both revenue and adjusted loss per share, though the stock closed the session lower before recovering modestly after hours. The quarter's defining event sat outside the quarter itself, with the $1.8 billion SkyWater acquisition closing on July 31 and creating what management describes as the first vertically integrated full-stack quantum platform. Full details are in the Q2 2026 earnings release.
Analyst Take
IonQ closed the June quarter as a different company than the one that opened it. The SkyWater transaction, valued at $1.8 billion and completed after quarter end, hands IonQ ownership of the fab that builds its ion trap chips. Vertical integration in deep tech usually reads as a de-risking move, so it is no surprise here. It also introduces a wrinkle that received less attention on the call than we think it warrants. Management disclosed that full-year IonQ spending with SkyWater under their commercial agreement would have run to roughly $120 million, money that converted into SkyWater revenue and now eliminates on consolidation. The bull case is straightforward and largely correct. Eliminations are an accounting artifact, cash outflow is unchanged, and owning the iteration cycle is worth more than a tidy revenue line. Our concern is narrower, and it has a deadline attached. SkyWater files its own second quarter report as a separate registrant covering a period in which IonQ was, by management's own description, its largest customer. That filing is the last clean external read on the foundry's standalone commercial pull before consolidation folds it into a single line.
What The Numbers Mean
The outperformance was deployment-timed rather than subscription-timed, and that distinction matters for how the second half should be modeled. Management attributed the largest share of the beat to fifth-generation system shipments, specifically subsystems moving to KISTI in Korea and a machine entering final assembly at QuantumBasel in Switzerland, installed alongside the prior-generation system that customer already owned. Consecutive generations sitting side by side at one commercial site is a useful proof point for a company that has spent years arguing it can manufacture at cadence rather than build one-offs. It is also lumpy revenue by construction. Management drew RPOs down into revenue during the quarter and replenished them past the starting balance, which is the more durable signal.
The spending story is where the print gets less comfortable. Adjusted EBITDA loss widened materially, and management attributed roughly $20 million of that to incremental SkyWater spend as the technology roadmap accelerated, plus approximately $10 million of pre-integration investment ahead of the close. That is a company paying twice, once to fund an external partner and once to prepare to absorb it. The GAAP net loss is largely noise, driven by warrant remeasurement rather than operations, and management was right to say so plainly.
Two disclosures deserve tracking. QCaaS is reportedly growing faster than organic computing revenue and carries higher margin, yet remains unquantified. And the Anduril and Sandia memoranda carry no dollars in RPO. Both were framed as upside. Neither is in the numbers yet.
Market Analysis
The same-quarter prints across listed quantum names describe a sector separating into two tiers rather than converging. Rigetti reported second quarter revenue of $5.1 million, up sharply year over year on on-premises Novera QPU deliveries, with improving gross margin. D-Wave reported $3.1 million, essentially flat year over year, with bookings and remaining performance obligations rising while recognized revenue did not. Both prints tell the same structural story IonQ's does. Revenue recognition in this sector is governed by system delivery timing, not by demand.
Capital formation is where the peer set genuinely diverges. Rigetti's quarter carried a Department of Commerce letter of intent for CHIPS-linked funding, and the letter contemplates an equity stake for the Department in exchange. That is a second model for how Washington capitalizes this sector, and it sits in direct contrast to IonQ, which funded a fab acquisition off its own balance sheet and took on no federal ownership to do it. Whether investors eventually prefer sovereign-backed dilution or self-funded dilution is an open question. They are not the same trade.
Two upcoming prints should sharpen the picture inside a single week. Quantum Computing Inc. reports on August 10, which will indicate whether photonics foundry economics follow the same delivery-lumpy pattern. Quantinuum reports on August 11 in its first quarter as a public company, and that print matters more. Quantinuum is the closest architectural comparable to IonQ, trapped ion and full stack, and its March quarter revenue fell to $5.2 million from $19.1 million a year earlier, with customer concentration swinging hard between periods. If Quantinuum's June quarter shows the step function IonQ posted, the sector narrative shifts toward a broader demand cycle. If it does not, IonQ's separation looks company-specific.
The Error Correction Question
The more interesting convergence this quarter is not commercial. It is architectural. IonQ reported breakeven quantum error correction using qLDPC codes on a Tempo engineering test system, validating elements of its Walking Cat blueprint. That result does not stand alone. IBM's Loon architecture introduced c-couplers to enable long-range connections between distant qubits on a single chip, specifically to support high-rate qLDPC codes, and IBM has demonstrated real-time qLDPC decoding on classical hardware in under 480 nanoseconds. IQM announced its own qLDPC family in June, engineered to avoid long-range crossing couplers and simplify fabrication. Riverlane forecast exactly this pattern, predicting that other industry players would follow IBM's move to qLDPC codes in 2026, producing diverse fault-tolerant architectures tailored to specific hardware platforms.
So while the codes are converging, the hardware is not. That appears deliberate.
Three distinct bets are now visible on where error correction should physically live. IBM is betting on the code layer, buying overhead reduction with chip connectivity. HRL is betting on the control layer, publishing a silicon spin processor whose cryogenic CMOS controller ran error correction autonomously, with roughly fivefold suppression as qubits were added. D-Wave, through its Quantum Circuits acquisition, is betting on the qubit itself, with peer-reviewed dual-rail erasure results showing a two-qubit entangling gate at roughly 99.9% accuracy with native hardware-level error detection. IonQ is arguably running two of the three, pairing qLDPC codes with the electronic qubit control it acquired from Oxford Ionics.
The unifying constraint is manufacturing, not physics. The McKinsey Quantum Technology Monitor for 2026 identified manufacturing systems and cryogenic infrastructure, rather than raw qubit count or error rate, as the field's binding constraint. Read the consolidation wave against that finding and it stops looking opportunistic. IonQ bought a fab. IBM moved production to 300mm. HRL built its controller on commercial CMOS lines. D-Wave bought a qubit architecture. These are supply chain decisions wearing physics clothing.
Two cautions. Breakeven error correction on a test system is a long way from fault tolerance in a commissioned product. Convergence on method is not convergence on schedule. Also, Overhead ratios are not measured consistently across platforms, so cross-vendor comparisons of physical-to-logical efficiency should be treated as directional. Assessing this sector places industry analysts in an exceptionally difficult position. Their institutional toolsets are legacy enterprise IT evaluation frameworks, designed for mature software packaging, predictable SaaS deployment, and linear go-to-market execution. There is a real challenge when applying these tools with an enormous amount of brand identification to a deeply hybridized, high-performance computing frontier whose core physics remain actively contested. When analyst methodologies require squeezing a fluid, hardware-bound paradigm into a rigid 2-dimensional grid, the scoring naturally favors near-term middleware polish and short-term customer onboarding. These factors may thus ignore the impact of foundational breakthroughs in gate fidelity, logical scaling, and error correction. It is less a reflection of analyst acumen and more a structural mismatch: an institutional insistence on treating an evolving HPC physics trade as if it were just another enterprise software category.
Looking Ahead
Based on what we are observing, the IonQ September 8 Investor Day carries more analytical weight than this print did. Management declined to guide the combined company, citing a close that landed days after the call, purchase price accounting and intercompany eliminations still to be sized. That is a defensible position and an honest one. It also means the market is currently valuing a company whose consolidated revenue base has not been disclosed in any form.
The demand side, meanwhile, now has dates attached to it. Executive Order 14412 sets a December 31, 2030 deadline for federal agencies to move their most sensitive systems to post-quantum encryption and a December 31, 2031 deadline for post-quantum authentication. The EO also directs federal contractors to comply with post-quantum FIPS standards by the end of 2030. Companion order 14413 addresses computing, sensing, networking and supply chain. That converts IonQ's security portfolio from a thesis into a procurement line item. Independent corroboration matters here, and it exists because Cloudflare moved its own full post-quantum target to 2029 in April 2026, citing research breakthroughs. Thats a notable move because Cloudflare has no quantum hardware to sell. The variable we will track most closely is whether 256-qubit commissioning holds to the first half of 2027, because that date is now the hinge on which the semiconductor thesis turns.
Stephen Sopko | Analyst-in-Residence – Semiconductors & Deep Tech
Stephen Sopko is an Analyst-in-Residence specializing in semiconductors and the deep technologies powering today’s innovation ecosystem. With decades of executive experience spanning Fortune 100, government, and startups, he provides actionable insights by connecting market trends and cutting-edge technologies to business outcomes.
Stephen’s expertise in analyzing the entire buyer’s journey, from technology acquisition to implementation, was refined during his tenure as co-founder and COO of Palisade Compliance, where he helped Fortune 500 clients optimize technology investments. His ability to identify opportunities at the intersection of semiconductors, emerging technologies, and enterprise needs makes him a sought-after advisor to stakeholders navigating complex decisions.