1
1TL;DR: IonQ just had its best quarter ever—revenue nearly quadrupled year-over-year. They also bought a semiconductor factory for $1.8 billion to build their own quantum chips. But they’re still losing lots of money on paper due to accounting rules.
| Metric | Q2 2026 | Change | Why It Matters |
|---|---|---|---|
| Revenue | $80.1 million | +287% YoY | Massive growth, beat expectations by 20% |
| Full-Year Guidance | $280–290M (standalone) | Raised | Confidence in continued momentum |
| Organic Growth | ~100% expected | Maintained | Core business doubling yearly |
| GAAP Net Loss | $1.9 billion | Mostly non-cash | Accounting quirk, not cash burn |
| Adjusted EBITDA | -$120.3 million | Investment phase | Spending to build future capacity |
| Backlog (RPO) | $485 million | Up from $122M YoY | Future revenue pipeline growing |
IonQ (NYSE: IONQ) reported $80.1 million in Q2 2026 revenue—that’s 287% more than the same quarter last year. CEO Niccolo de Masi called it their "strongest quarter to date" and the fifth straight quarter of record results.
ELI5: Imagine a lemonade stand that made $100 last summer. This summer, it made $387. That’s what 287% growth looks like.
COO/CFO Inder Singh said revenue came in 20% above their internal expectations. The company also raised full-year guidance to $280–290 million (for IonQ standalone, not including the new acquisition).
IonQ’s customer base is becoming impressively diverse:
ELI5: "Remaining Performance Obligations" is fancy accounting talk for "customers have signed contracts promising to pay us this much in the future." It’s like having a stack of signed IOUs.
IonQ began shipping subsystems to the Korea Institute of Science and Technology Information (KISTI). Equipment is being delivered and assembled on-site.
At QuantumBasel, a 5th-gen system is in final assembly right next to a previously purchased 4th-gen system. Singh called this "the first commercial deployment of two successive generations of quantum computers at the same customer site."
Why this matters: It’s like a customer buying both the iPhone 15 and iPhone 16 to run side-by-side. Shows deep trust and expanding use cases.
A U.S.-based semiconductor foundry—a factory that designs, fabricates, and packages chips for other companies. Think of it as a "chip factory for hire."
| Reason | Explanation |
|---|---|
| Control their destiny | No more waiting on outside factories |
| Trusted U.S. facilities | Critical for government/defense customers |
| Vertical integration | Design → Fab → Package → Deploy all in-house |
| Scale manufacturing | Use standard semiconductor processes to mass-produce quantum chips |
IonQ is transitioning from laser-based qubit control (bulky, hard to scale) to electronic qubit control (chip-scale, manufacturable). This technology came via their Oxford Ionics acquisition.
ELI5: Old way = controlling each quantum bit with a separate laser beam (like conducting an orchestra with 100 laser pointers). New way = controlling them with tiny electrical signals on a chip (like a modern smartphone processor).
ELI5: "Fault-tolerant" means the computer can fix its own mistakes as it runs. Current quantum computers are like a calculator that randomly forgets what 2+2 equals. Fault-tolerant ones catch and correct those errors automatically.
IonQ isn’t just building computers—they’re building security for the quantum age.
IMPORTANT CALLOUT: Don’t Panic About This Number
| Here’s the breakdown: | Component | Amount | Reality Check |
|---|---|---|---|
| Non-cash warrant revaluation | ~$1.6 billion | Pure accounting—no cash left the building | |
| Actual operating loss | ~$300 million | Still high, but investment-phase normal | |
| Adjusted EBITDA | -$120.3 million | Better measure of cash burn |
Warrants are like stock options for investors. When IonQ’s stock price goes up, accounting rules say the liability of those warrants goes up—creating a "loss" on paper.
ELI5: Imagine you promised your friend "I’ll give you $10 if my baseball card collection doubles in value." If your cards suddenly become worth $1,000, accounting says you now "owe" more—even though you didn’t spend a dime. That’s a non-cash mark-to-market loss.
IonQ will host an Investor Day at the New York Stock Exchange on September 8, 2026. Expect deep dives on:
| What’s Going Right | What to Watch |
|---|---|
| Revenue nearly 4x YoY | Still burning cash (~$120M/quarter adjusted) |
| Diverse, global customers | No combined guidance yet post-SkyWater |
| Own chip factory now | Integration risk with SkyWater |
| Clear path to 256 → 10K qubits | Warrant accounting noise will continue |
| Security business growing | Competition (IBM, Google, Rigetti, PsiQuantum, etc.) |
| $485M backlog = visibility | Tech risk: trapped-ion vs. superconducting vs. photonic |
Bottom Line: IonQ is executing aggressively on a vertical integration strategy—buying the factory, inventing the chip architecture, and selling the full stack (hardware + cloud access + security + algorithms). They’re in the "invest heavily now, profit later" phase. The revenue growth proves customers are buying. The losses prove they’re building.
No. They lost $1.9B on paper (mostly accounting) and burn ~$120M/quarter in real cash (adjusted EBITDA). They’re in growth/investment mode.
Q2 2025 revenue was ~$20.7M. Q2 2026 was $80.1M. That’s nearly 4x the revenue in one year.
Control, speed, and IP protection. Owning the fab means IonQ can:
Qubits = quantum bits. More qubits = exponentially more computing power (in theory). But quality matters more than quantity—error rates, connectivity, and coherence time are critical. IonQ’s "walking cat" architecture aims for high-quality, error-corrected qubits at scale.
I can’t give financial advice. But consider: high growth + high burn + binary technology risk + accounting volatility = very high risk/reward. Read the 10-Q, watch the Sept 8 Investor Day, and decide based on your risk tolerance.
Article based on IonQ Q2 2026 earnings release and conference call highlights. Original reporting by MarketBeat. Not financial advice.