◆ Inside the Latest Results (2025 & Q2 2026)

Bloom's recent results are the financial record of one of the most dramatic turns in the market: a company that lost money for two decades becoming, almost overnight, a profitable and explosively-growing beneficiary of the AI-power boom. Fiscal 2025 was already a step up — revenue rose 37% to $2.02 billion¹ — but the company still posted a small net loss, the twentieth-odd consecutive year of red ink, and the story looked like the same slow grind it had always been. Then, in 2026, the AI-data-center demand surge hit the financials with full force, and everything changed at once.

The second quarter of 2026 was the inflection made visible: revenue reached a record $1.065 billion, up 166% year-over-year², the company's first billion-dollar quarter — driven by product revenue that leapt 215% to $935 million as hyperscalers, neoclouds, and colocation providers bought Energy Servers to power AI data centers the grid could not serve fast enough. Non-GAAP gross margin expanded over six percentage points to 34.3%, and non-GAAP earnings of $0.78³ a share blew past the roughly $0.40 expected. On the strength of it, Bloom raised full-year 2026 guidance substantially — to $3.9–4.2 billion of revenue (from $3.4–3.8 billion), $800–900 million of operating income⁴, and $2.55–2.85 of non-GAAP EPS — and pointed to a total backlog of some $20 billion and a financing partnership with Brookfield expanded from $5 billion to $25 billion to fund the deployments. A perpetual money-loser had, in a single boom, become a profitable company guiding to nearly a billion dollars of operating income.

The results are genuinely transformative and should not be dismissed: they represent the moment Bloom's twenty-year effort to make the economics work finally succeeded, driven by a real and enormous source of demand, with contracted backlog and committed financing lending it substance. This is the evidence for the bull case, and it is real evidence — not a narrative but a billion-dollar quarter and a raised profit guide. An investor should credit the achievement fully.

But the honest reading holds two cautions alongside the celebration. First, these are, by their nature, peak results — the product of an extraordinary, possibly-temporary demand surge, with margins inflated by the speed premium of scarce fast power, and profitability that is only a few quarters old and untested by any normalization. Second, the market's own reaction was telling: the stock gained and then gave back a double-digit move around the report, a sign that even blockbuster results struggle to satisfy a valuation that has already priced in extraordinary success. Read correctly, the latest results tell a consistent story: Bloom's business has genuinely and dramatically improved, the AI-power demand is real and large, and the company has finally turned profitable at scale — and all of it has been achieved at the peak of a boom, priced as though the peak will persist, for a company whose moat remains thin. The numbers are a triumph and a warning at once: proof that the demand and the profits are real, and a reminder that they are the best of times for a business whose durability through the rest of the cycle is exactly what these spectacular figures cannot yet demonstrate.