IonQ published its second-quarter report on 5 August 2026. We are writing about it today, 6 August.
And it is a report where two apparently contradictory numbers appear in the same release, and both are correct.
What was reported
| Quarter | Change | |
|---|---|---|
| Revenue | 80.1 | +287% |
| Net loss | (1,867.7) | - |
| Loss per share | $(5.08) | - |
In millions of dollars except per-share data
Revenue of $80 million. A loss of $1.87 billion. Twenty-three times over.
The explanation, and it is almost entirely one line
$1.649 billion from revaluing warrants
Out of a $1,867.7 million loss, a single line accounts for $1,649.1 million:
Change in the fair value of warrant liabilities.
That is roughly 88% of the loss, and it involves no cash.
What happens there? The company has warrants that trade, and the accounting rule requires the liability for them to be remeasured each quarter at market value. When the share price rises sharply, the warrants become more valuable, the booked liability grows - and that is recorded as a loss.
In other words: the more the stock rises, the larger the accounting loss.
That is exactly the line we saw this week at TeraWulf - $755.7 million of a $940 million loss - and at Cipher, $150.5 million.
The loss is real in accounting terms. It simply does not describe the business.
And what does describe the business
Revenue: $80.1 million, up 287%.
And more than that - it came in 20% above the midpoint of the range the company itself had guided.
The revenue split the company discloses:
- Roughly 50% international
- Roughly 60% commercial - meaning non-government
- Roughly 25% multi-product
And the figure I consider the most important: remaining performance obligations grew 297% year over year.
That is the contracted backlog - work signed but not yet recognised as revenue. When it grows at a rate similar to revenue, it means the growth is not draining the backlog but replenishing it.
The guidance
The company raised full-year guidance to $280 to $290 million, and stated confidence in achieving 100% organic growth for full year 2026.
And what deserves qualifying
First, the comparison base. Growth of 287% comes off a base of roughly $20 million a quarter. That is easier than growing 30% off a base of a billion.
Second, share-based compensation. The cash flow statement shows share-based compensation of $270.4 million, against $132.4 million a year ago. That is more than three times the quarter's revenue - and it is a real cost to the shareholder, paid in dilution.
And third, the SkyWater acquisition. The company closed it on 31 July, after quarter end, and states explicitly that the results and outlook do not include it. So the picture changes next quarter, and not only from the existing business.
And the detail that connects to a wider thesis
The company reports that its on-orbit optical communications terminals reached a record 84, supporting a U.S. Government initiative.
That is not quantum computing - it is optical communications. And IonQ is building a second business there, alongside the quantum computer itself.
And it touches exactly the place Tower touches with its silicon photonics, and Arista with the optics saving 60% of interconnect power. Moving information as light rather than electricity is a thesis running across several of the companies we covered this week.
My Angle
A personal opinion of Ilan Abramov - not advice, not a recommendation
This is a report you cannot read without taking it apart, and I think that is exactly what makes it interesting.
Anyone seeing a $1.87 billion loss on $80 million of revenue concludes the company is heading for collapse. And anyone seeing 287% growth concludes it is heading for the moon. Both are reading the same report, and both are wrong.
The number I look at is remaining performance obligations, up 297%. Revenue growth can be manufactured in a single quarter. A contracted backlog growing at the same rate says there is more signed work behind it - and that is what separates a jump from a trend.
And what I put a large question mark on is share-based compensation: $270 million, more than three times revenue. That is not a theoretical accounting charge - it is a transfer of ownership from existing shareholders to employees. At this stage of a company it is customary, but at this scale it stops being a technical detail.
And what I am waiting for is the next quarter, because of SkyWater. An acquisition closed after quarter end and excluded from guidance means the numbers we are looking at today no longer describe the company that will report next time.
