Aehr Test Systems: Wafer-Level Burn-in Takes Flight as AI Demand Reshapes the Semiconductor Test Landscape
The Inflection
Aehr Test Systems' fiscal Q4 2026 report was anything but routine. The company posted record quarterly bookings of $60.7 million—up over 500% year-over-year—and ended the year with a record backlog of $80.6 million, later swelling to an effective ~$100.6 million with post-period orders. CEO Gayn Erickson opened the call with a clear message: “we are expecting revenue of between $130 million and $150 million representing 2.6x to 3x the just-completed fiscal 26 revenue.” — Gayn Erickson · 2026-07-14 This is not a continuation of the recent trend; it is a step-change in scale.
The Pivot
The company's transformation over the past two years is stark. As Erickson noted, “Just 2 years ago, over 95% of our business was tied to silicon carbide for electric vehicles. Whereas today, almost 95% of our fiscal year 2020 revenue came from markets other than electric vehicles silicon carbide.” The keyword trajectory captures this pivot: wafer level burn in system has surged to the top of the company's quarterly rankings, while AI data center has become a recurring theme. The bulk of the fiscal 2027 guidance is already covered by existing customers, and the company explicitly excludes any memory revenue from its base case, signaling that the upside is larger than the headline number suggests.
Wafer-Level Breakthrough
The most compelling evidence of the inflection came from the benchmark with a major AI processor supplier. Management reported that the wafer-level burn-in solution exceeded expectations, with the customer explicitly stating it produced results better than they can get at the package level.
This has moved the customer to consider pilot production validation for its current high-volume device, and the company sees a significant revenue opportunity. As Erickson put it in Q&A, “the potential revenue opportunity from 1 of these devices is significant to Aehr in terms of near and long term revenue streams” — Gayn Erickson · 2026-07-14. The capacity need is real; customers are already asking for more systems and WaferPaks, and the company is scaling manufacturing capacity through a Southeast Asian contract manufacturer to meet the anticipated demand.We are excited to report that we successfully completed the benchmark testing of our wafer-level burn-in solution on 1 of their processors. Achieving results that exceeded their expectations and, in their words, produced results better than they can get at the package level.
Capacity and Risk
Execution remains the key question. The company has expanded its contract manufacturing capacity, with the ability to ship over 20 Sonoma systems per month, but supply chain constraints—particularly power supplies—are a risk. The company also faces the challenge of an acyclical lumpy demand environment, as Erickson candidly noted: “We know how to be lumpy.” The diversification into memory (HBM/flash) is a future opportunity, with orders possible in fiscal 2027 and revenue in 2028, but management remains appropriately cautious. The prior quarter's language on the TAM was optimistic: “The estimates of -- today, if you were to look at AI spend in test between test and burn-in, is it $8 billion, $10 billion to maybe $15 billion or so.” — Gayn Erickson, President and Chief Executive Officer · 2026-01-08 The potential is clear, but so is the uncertainty.
Valuation Perspective
Investors have already re-priced the stock: it rose over 44% in the past 90 days, reaching an all-time high of $145.61 on August 17, before pulling back over 30% into the report. The fundamentals are still in transition—revenue declined 15% in fiscal 2026 and gross margins have compressed. Yet the balance sheet is strong, with effective net cash of $32 million after raising ~$100 million via an ATM. At a price-to-sales multiple above 25x, the market is paying for the inflection, not the trailing results. The record backlog and the validation from a top-tier AI processor supplier make the risk worthwhile, but the execution over the next two quarters will be critical to justify the multiple.
In summary, Aehr Test Systems has moved from a niche silicon carbide tester to the center of the AI-driven burn-in wave. The second device qualification with a leading hyperscaler, the benchmark success, and the record backlog provide concrete evidence that the story is real. The market has taken notice; the question now is whether the company can deliver on its audacious outlook.