Aehr Test Systems' Q1 2026 Earnings Call: Contradictions on Ai Market Potential, Wafer-Level Burn-In, and Flash Memory Growth

The above is the analysis of the conflicting points in this earnings call

Date of Call: October 6, 2025

Financials Results

• Revenue: $11.0M, down $2.1M YOY vs $13.1M prior year
• EPS: $0.01 non-GAAP diluted EPS, down from $0.07 in the prior year
• Gross Margin: 37.5% non-GAAP, down from 54.7% in the prior year

Business Commentary:

* Revenue and Financial Performance: - Aehr Test Systems reported Q1 revenue of $11 million, a $2.1 million decrease from $13.1 million in the same period last year. - The decline in revenue was mainly due to lower sales volume and a less favorable product mix compared to the previous year, which included a higher volume of higher-margin WaferPaks.

• Growth in Wafer-Level Test and Burn-in:
• The company shipped multiple FOX-CP single wafer production test and burn-in systems for new high-volume applications involving burn-in and stabilization of new devices for a leading hard disk drive customer.

• This growth was driven by demand for wafer-level test and burn-in solutions in the hard disk drive market and ongoing opportunities in other segments like silicon photonics and gallium nitride semiconductors.

• AI and Wafer-Level Burn-in Opportunities:

• Aehr Test Systems saw momentum in the qualification and production burn-in of packaged parts for AI processors, leading to follow-on volume production orders for their Sonoma systems from a leading hyperscaler.

• The interest in wafer-level burn-in for AI processors was underscored by a paid evaluation program with a top-tier AI processor supplier for full wafer-level test and burn-in capabilities, indicating a promising market for advanced test solutions.

• Facility Expansion and Manufacturing Capacity:

• The company completed a $6.3 million renovation of their manufacturing facility, significantly upgrading the manufacturing floor, test labs, and clean room space for high-power systems.
• This expansion increased overall manufacturing capacity by at least 5x, enabling the production of more high-power systems for AI configuration and supporting the growth of other markets.

Sentiment Analysis:

• Management said results exceeded analyst expectations for both revenue and profit. However, non-GAAP gross margin fell to 37.5% from 54.7% YOY, and they are withholding formal guidance due to tariff-related uncertainty. They expressed confidence in broad-based growth opportunities (AI, silicon photonics, HDD, GaN, SiC) and cited an expanding backlog ($15.5M at Q1-end; $17.5M after first 5 weeks of Q2), but timing of larger AI orders remains uncertain and likely second half weighted.

Q&A:

• Question from Christian Schwab (Craig-Hallum Capital Group): When will bookings materially improve to drive revenue later in the year?
  Response: Expect follow-on capacity orders from the first AI wafer-level customer this fiscal year—potentially exceeding last year—with timing likely in the second half; no orders in hand yet.

• Question from Christian Schwab (Craig-Hallum Capital Group): How many AI customers do you aim to be shipping to by fiscal year-end?
  Response: Internal targets call for multiple new AI customers across both package-level and wafer-level; pacing depends on evaluations, but engagement is broad and growing.

• Question from Christian Schwab (Craig-Hallum Capital Group): On AI TAM being 3–5x SiC, will we see material orders this fiscal year?
  Response: Management is confident large orders will come as evaluations complete, with some programs targeting volume in H2 calendar 2026 requiring tools ahead of that.

• Question from Mark Shooter (William Blair): Do AI customer qualifications require a new product cycle, and what is customer risk appetite?
  Response: No new platform is needed; Aehr adapts WaferPaks and DFT to customer specs, with timing typically aligned to new product ramps and growing willingness to adopt given OSAT demonstrations.

• Question from Mark Shooter (William Blair): Why start with Sonoma (package) before wafer-level burn-in, and how does the transition occur?
  Response: Aehr supports both; customers often qualify and start production on Sonoma, then move to wafer-level to avoid costly multi-die package yield loss—Aehr is first to ship AI wafer-level burn-in.

• Question from Bradford Ferguson (Halter Ferguson Financial): Cost/efficiency of system-level burn-in versus wafer-level, and implications for HBM/HBF?
  Response: Wafer-level burn-in enables high-temperature accelerated stress at far lower power than system-level and is better suited for screening; both HBM and emerging HBF will require robust burn-in.

• Question from Larry Chlebina (Chlebina Capital Management): Does AMD’s OpenAI news accelerate your second AI processor evaluation?
  Response: They won’t name the customer, but they’re engaged with all major AI suppliers; broader processor momentum is positive and Aehr offers both package and wafer-level solutions.

• Question from Larry Chlebina (Chlebina Capital Management): For optical I/O, will demand be upgrades or new machines?
  Response: Both—customers are purchasing additional systems and upgrading existing ones, including integrated automation.

• Question from Larry Chlebina (Chlebina Capital Management): Is the HBF opportunity with the same flash customer, and what about the original enterprise flash program?
  Response: Same customer with evolving requirements; HBF supersets prior needs and legacy tools won’t suffice—Aehr is proposing configurations that can address both.