Atomera Incorporated (ATOM) Earnings

Management highlights

### Gate All Around (GAA) Advanced Logic - Atomera continues engagement with 2 of the 4 target GAA customers, and passed a significant manufacturing milestone with one customer that answered critical manufacturability questions about MST (Mears Silicon Technology) integration into unique GAA silicon structures. - The milestone is a required technical demonstration step before customer acceptance of new materials into production process flows; active discussions remain ongoing with the other 2 target GAA customers, supported by a strategic partner that provides test infrastructure and ecosystem credibility. ### Memory Market Expansion - DRAM customer interest shifted from planar periphery enhancements to vertical scaling architectures (4F² and 3D DRAM) driven by accelerated AI demand. Atomera has developed a new validated value proposition for 4F² DRAM that solves fundamental device challenges and reduces manufacturing costs via MST's precise doping profile control, with feasibility proven via TCAD simulation that will be presented at an IEEE conference in September 2026. - A new opportunity for planar periphery enhancements has opened in NAND flash memory, a market Atomera previously did not address. AI-driven demand for higher NAND performance has created a need for the MST technology Atomera developed for DRAM planar periphery; adoption by NAND manufacturers would more than double Atomera's total addressable market (TAM) for MST. ### Power and RF Development Pipeline - The large IDM customer development program is progressing on plan, with new device test data arriving as the next round of experiments begins. - Power semiconductor trench FET and HBT development continues advancing, targeted at the efficiency and high-frequency requirements of AI data centers. - RFSOI joint development with the second JDA partner is ongoing, and internal development of a high-throughput manufacturing process for RFSOI (focused on critical substrate supply chain requirements) is progressing well, with potential applicability to multiple other use cases. ### GaN on Silicon RF Breakthrough - Atomera achieved a major technical breakthrough for GaN on silicon RF devices, solving the existing parasitic channel problem with MST. Third-party testing by partner InSize delivered outstanding RF performance: linearity is roughly 1,000 times better than baseline GaN on silicon at benchmark drive levels, and remains 2-3 orders of magnitude better across the full power sweep, approaching the performance of expensive advanced trap-rich RF SOI. - The breakthrough was well received at the June 2026 International Microwave Symposium, and has generated new customer engagements for evaluation of MST-enabled GaN on silicon. - Strategically, this breakthrough could catalyze a shift from traditional RFSOI to low-cost GaN on silicon for RF front-end designs, enabling fully integrated front ends that include power amplifiers (which RFSOI cannot support), positioning Atomera at the start of a potential new high-growth market with MST enabling either development path customers choose.

Guidance

- Management previously guided 2026 full-year non-GAAP operating expenses to approximately $18.5 million, with a budgeted range of +/- $250,000 around this figure. - Due to industry-wide cost increases for outsourced engineering services, tool leases, metrology, and device fabrication driven by rapid semiconductor industry growth, management now expects 2026 full-year non-GAAP operating expenses to come in at the high end of the guided range, i.e. ~$18.75 million, an upward revision within the original guided range. No guidance for 2027 full-year results was provided, though management noted that structural cost increases (including upcoming tool lease renewals) will be visible in 2027 P&L results.

Segment performance

Atomera is an early-stage semiconductor technology firm, and all Q2 2026 revenue came from customer wafer delivery fees, primarily to the large IDM customer program. Total Q2 2026 revenue was $158,000 (100% of revenue). No separate product segments with distinct revenue figures are reported, as all current activity is pre-commercialization development focused on technology validation for future licensing and wafer sales.

Risks & headwinds

- All commercialization milestones have uncertain timelines and outcomes: technical demonstration milestones for GAA, DRAM, and GaN on silicon do not guarantee customer adoption, and could require multiple additional rounds of testing before licensing negotiations can conclude. - Atomera is facing significant cost increases from external service providers, including outsourced engineering, metrology, device fabrication, and tool leases. A major tool lease renewal is still in negotiation, with a material expected cost increase that will impact 2027 expenses. - Atomera did not win a proposed GaN power device development program through Power America, though management noted this program was not financially material. - Commercial adoption of new semiconductor materials is a long, multi-stage process, with no guarantee of successful commercial agreement even after technical milestones are passed.

Analyst Q&A

• Q: Can you explain the importance of the recent GAA milestone, what the next steps are, and the timeline and likelihood of eventual success? /

  A: The milestone cleared was a required demonstration that MST can be successfully deposited on the customer's complex GAA structure, answering their key manufacturability questions. A few more demonstration steps remain; after successful completion of all demonstrations, the customer will need to license MST to test it in their own fab (to protect their internal IP). Discussions of license terms have been ongoing since the start of the engagement, and will accelerate as milestones are completed. If all goes well, the move to licensing could happen as early as late 2026, but it will not occur in the next two months, and success is not guaranteed.

• Q: Has the DRAM market shift to vertical scaling reset the customer development cycle, and is the existing planar periphery technology now applicable to NAND? /

  A: Yes, the focus has shifted from our original planar periphery opportunity for DRAM to a new value proposition for 4F² vertical DRAM, restarting the evaluation and testing cycle with customers, who have already validated the concept. The planar periphery technology we spent 5 years developing for DRAM is now in high demand from NAND flash manufacturers due to AI-driven performance requirements, opening an entirely new TAM for Atomera that we never previously addressed, which is a major new opportunity.

• Q: How does your GaN on silicon breakthrough enable adoption of GaN for RF applications, and what does it mean for Atomera's TAM? /

  A: RFSOI is hitting performance limits for high frequency and high power applications, and GaN offers much better performance, but GaN on silicon carbide (the current mainstream solution) is too expensive for mass-market mobile RF front ends. GaN on silicon is low-cost, but previously suffered from poor RF performance that MST has now solved. MST-enabled GaN on silicon can support fully integrated RF front ends including power amplifiers (which RFSOI cannot do), which represents a major TAM expansion for Atomera if this design shift occurs, which Atomera is well positioned to enable via technology licensing.

• Q: What is the update on the Power America partnership, and what is your cash outlook? /

  A: Atomera did not receive the planned Power America GaN power development award; the program was not financially material (it would have generated only ~$300,000 in revenue), and the engineering resources that would have been allocated to it can now focus on the higher-priority GaN RF opportunity. Atomera ended Q2 2026 with $38.4 million in cash, equivalents, and short-term investments, which management believes is sufficient to execute on current opportunities, and the firm will remain disciplined on cost controls.