Atomera (ATOM) reported a third-quarter loss of $0.15 per share on no revenue. On a per-share basis, earnings improved 11.8% from the year-ago loss of $0.17, while revenue fell 100.0% from a token $11,000. That per-share improvement deserves scrutiny. In dollars, the net loss widened to $6.0 million from $5.6 million a year ago. Weighted shares rose to 38.9 million from 31.1 million, and that dilution, not better economics, produced the better EPS. The central tension is that Atomera had arguably its best quarter of commercial-pathway building, adding a third gate-all-around customer and signing a GaN joint development agreement with IQE. Yet it booked zero revenue and still has no signed license with a GAA, memory, or analog customer.
The quarter looks better against the prior quarter than against last year. Total operating expenses fell to $6.3 million from $6.9 million in the June quarter, driven by a drop in general and administrative costs to $2.3 million from $3.2 million. Research and development rose to $3.6 million from $3.3 million, which is the right direction for a company trying to convert engagements into deals. The adjusted EBITDA loss improved to $4.6 million from $5.0 million sequentially, though it was slightly worse than the $4.4 million loss a year ago. Per the call, operating cash use fell for a second straight quarter, to $3.2 million from $3.9 million in Q2 and $4.6 million in Q1. Atomera also made no at-the-market share sales for a second consecutive quarter. Cash and short-term investments stood at $34.8 million, up from $19.2 million at year-end after earlier equity raises but down from $38.4 million at the end of June.
The operating story is about engagement depth. Atomera is now working with three of the four gate-all-around foundry targets, up from two in Q1 and Q2 and one at the end of 2025, which meets the objective management set in February. Two of those customers are running MST on their own proprietary silicon, and management said every milestone set so far has been met. Memory work broadened from a 4F² DRAM concept last quarter to DRAM and NAND engagements, supported by a peer-reviewed paper presented in September. Under the new 'Prima' branding, starting-wafer applications grew from two to six over the year. Management expects Prima products to reach revenue faster than the in-device 'Selecta' line.
The IQE agreement is the most tangible new catalyst. It moves GaN-on-silicon from lab results earlier in the year to a production-oriented partnership covering RF and power applications. Management described it as a short development effort, with qualification potentially only months after tuning. Management's tone toward STMicro also firmed, with the CEO saying 'eventually we will do a deal together.' That is encouraging, but it is not a contract.
Bears have legitimate ammunition. Management guided fourth-quarter revenue to only about $50,000. Back in February, Atomera expected commercial deals in 2026. With one quarter left, no GAA license or JDA has been signed, and production timelines remain unclear. Memory efforts are still described as early stage. Costs are also an issue. The 2026 non-GAAP opex guide was held at the high end of $18.5 million plus or minus $0.25 million. Industry inflation in tool leases, metrology, and fabrication persists, and management withheld 2027 opex guidance pending lease negotiations. A roughly $1.4 million one-time prepayment for 2027 tool leases will raise fourth-quarter cash burn, partly offsetting the recent improvement.
From a market perspective, the stock opened at $5.14 after the August report, already below its 200-day moving average of $5.48. That suggests the market had been treating engagement news with skepticism rather than paying up in advance. For a pre-revenue licensing company, the share price tends to hinge on contract announcements rather than quarterly losses. The next signed agreement matters far more than this quarter's EPS. The balance sheet buys time: at the Q3 operating burn pace, current cash covers well over two years, although rising lease and tooling costs could shorten that.
The bottom line is that Atomera's third quarter improved the quality of its pipeline without improving its income statement. The third GAA engagement, the IQE JDA, and broader memory work make the commercialization story more credible than it was a quarter ago. The headline EPS gain, however, reflects dilution rather than operating progress. Zero revenue, a minimal Q4 guide, withheld 2027 cost guidance, and a still-unfulfilled promise of 2026 deals mean the burden of proof has not shifted. Investors now need signatures, not just engagements.