IonQ reported earnings on 5 August, and the headline number was a genuine cracker: revenue of $80.1m, up 287% on a year earlier, a fifth straight record quarter and a comfortable beat. Then you reach the bottom line, a GAAP net loss of $1.9bn, and you might spit out your tea. Relax. Most of that is an accounting mirage. But the more interesting move happened five days earlier, and it tells you where quantum computing is actually going.
On 31 July, IonQ closed its $1.8bn purchase of SkyWater Technology, the largest pure-play semiconductor foundry (a factory that manufactures chips to order for other companies) based in the United States. The deal, announced back in January, was cleared by America’s Federal Trade Commission without any conditions. SkyWater keeps its name and its outside customers, but IonQ now owns the fab.
Why does a quantum computing firm want a chip factory? Vertical integration, the strategy of owning your whole supply chain rather than renting bits of it. IonQ builds trapped-ion machines, which hold individual charged atoms in place with electromagnetic fields and nudge them with lasers to do computation. Making the specialised chips and optics for that is fiddly, and leaning on outside suppliers is slow. Bring the fab in-house and, in theory, you control quality, cost and secrecy, and you iterate faster. IonQ says it now wants to test a 200,000-qubit processor in 2028.
Here is the angle: the quantum sector is moving from “look at our lab results” to “look at our factory”, and that shift matters more than any single earnings beat. Owning manufacturing is what grown-up hardware companies do. It is also expensive, capital-hungry and a long way from the tidy story of pure science. When a research darling buys an industrial fab, it is telling you it expects to make real machines at real volume, and soon.
Now, that $1.9bn loss. Most of it is a $1.6bn non-cash charge from warrant revaluation. A warrant is basically a coupon that lets someone buy shares later at a fixed price. When your share price rises, the accounting value of those outstanding coupons rises too, and you book it as a loss, even though no actual cash left the building. It makes the reported number look apocalyptic while the bank balance is fine. Worth understanding before you panic, or before someone sells you panic.
The sober bit: IonQ’s revenue, record or not, is still measured in tens of millions against a valuation in the tens of billions. This remains a story stock, beloved of retail traders, priced for a future that has not arrived. Buying a foundry makes that future feel more concrete. It does not make it certain. Quantum computing is still years from broad, profitable, error-corrected usefulness, and anyone telling you otherwise is selling something.
This is analysis, not investment advice. Quantum stocks are volatile. Do your own homework.
Did you know: the same regulator that has picked very public fights with the biggest names in tech waved through IonQ’s $1.8bn foundry deal with no remedies at all, an unusually clean run through the FTC in 2026.
Related on Top Tool Stack: NTT’s room-temp quantum bet · PsiQuantum’s $125m DARPA deal