Atomera's GaN and Memory Bets: Multiple Vectors, Same Enabler
GAA milestone, 4F2 DRAM traction, and a NAND TAM double — but the GaN RF data is the breakout.
ATOM · Earnings Call · 2026-08-04
A Quarter of Milestones
Atomera (ATOM) reported a Q2 that was less about the numbers (revenue of only $158K) and more about the narrative. The company, which licenses its MST (Mears Silicon Technology) to leading-edge semiconductor fabs, threaded together three separate strategic advances in a single quarter. First, in gate-all-around (GAA) logic, management disclosed that it had passed a key milestone with one of its two active customers. As CEO Scott Bibaud put it: “we've cleared a hurdle with our gate-all-around customers” — Scott Bibaud · 2026-08-04 — meaning MST deposition on the customer's own structure is now proven, pushing the engagement one step closer to an installation license. In the Q&A, he detailed that the next phase would require the customer to take MST into its own fab, a step that mandates a license agreement.Second, the memory story shifted. Earlier work had focused on planar periphery enhancements for DRAM, but rising AI demand has pushed DRAM makers toward 4F2 vertical (3D) architectures. Atomera pivoted to a new value proposition: using MST's doping control to simplify the 4F2 access transistor. Scott said: “"in 4F2, we have some very compelling technology where MST through our doping control capabilities can really help them to simplify their manufacturing process.” — Scott Bibaud · 2026-08-04 A TCAD simulation study will be presented at an IEEE conference in September, adding credibility.
Third, and most striking from a TAM perspective, the company announced a brand-new market: NAND flash. Historically, Atomera had no path into NAND because the planar periphery there was not a bottleneck. But AI-driven performance demands have changed that. "We have learned from a major NAND supplier that AI is now pushing NAND to the point that they need the planar periphery boost that MST can provide," Scott said. He later elaborated: “NAND actually manufactures more wafers per year than DRAM... For us, where we're selling products based on wafer shipments, that's a really good opportunity.” — Scott Bibaud · 2026-08-04 If adopted, this effectively doubles the addressable market for MST from memory.
The GaN RF Breakthrough
The most consequential news, however, was the GaN-on-silicon RF data. Atomera has long argued that MST could fix the parasitic channel problems that plague RF SOI-class performance on cheaper silicon substrates. This quarter, third-party RF tests from Incize confirmed that MST-enable GaN-on-silicon is now approaching the linearity and insertion loss of advanced trap-rich RF-SOI, while carrying GaN's inherent high-power headroom. That combination could tempt designers to swap RF-SOI switches and LNAs for GaN-on-silicon, and even integrate power amplifiers into the same front-end module — a shift that would open a large, new RF market. As Scott summarized in his prepared remarks:The company now has several new potential customers evaluating the technology in their own designs, a direct result of the IMS conference in June. Notably, this RF success also validates the earlier strategic pivot from GaN-on-power (where the company failed to win a PowerAmerica award) to GaN-on-RF — a pivot reinforced by the measured data.We believe some designs that would traditionally be built in RF-SOI could instead move to GaN-on-silicon. That would be a meaningful shift in how RF front-end designs get built. And MST's performance may well be the catalyst that sets it in motion. If new RF design activity begins migrating towards GaN-on-silicon, Atomera would be positioned right at the start of a new high-growth market, and it's worth underscoring that MST is the enabler on both sides of that shift. So whichever path the customer chooses, Atomera benefits.
This GaN narrative is a genuine company-unique signal, not sector boilerplate. It builds on prior commentary: in the May call Scott had noted “just looking at that data could be enough for someone to adopt us because it's such a big breakthrough” — Scott Bibaud, Chief Executive Officer · 2026-05-05 — a statement that now has hard numbers behind it. The same sense of escalating conviction applies to GAA, where in February he admitted “we've rarely been as excited about some technology results inside the company as we are by what we have right now” — Scott Bibaud, Chief Executive Officer · 2026-02-12 — and that enthusiasm has translated into a concrete milestone this quarter.
Financial Runway and Risks
Atomera ended Q2 with $38.4M in cash, providing roughly 8.9 quarters of runway at the current burn rate. Cash used in operations was $3.9M in the quarter, slightly better than Q1's $4.8M, but the start of multiple evaluation engagements and rising outsourced fabrication costs are expected to push 2026 non-GAAP operating expense to the top of the $18.25–$18.75M range. The company has not yet signed a single commercial license, and management remains deliberately cagey about timing, warning that a GAA license could slip by another nine months if a second demonstration round is needed.The stock price reflects this tension: after a spring rally that peaked at $11.17 in late May, ATOM has fallen ~46% over the last 12 weeks. The market is clearly pricing in execution risk, even as the technology blog posts and white papers accumulate. Still, the combination of a GAA milestone, a validated 4F2 DRAM concept, a brand-new NAND opportunity that doubles the TAM, and an RF GaN breakthrough with hard third-party data makes this quarter the most “catalytically dense” in Atomera's history. The financials remain pre-revenue, but the range of inflection points has expanded materially. If any one of these vectors converts into a license, the shares — still down ~90% from their 2021 peak — would likely re-rate sharply. The go-forward question is not whether the technology works (the data increasingly says it does), but whether the industry's conservative adoption cycles finally intersect with a hungry, well-funded startup.