Aehr Test Systems Q4 2026 Earnings Call Summary

Aehr Test Systems Q4 2026 Earnings Call Summary Moby Intelligence Wed, July 15, 2026 at 2:18 AM GMT+2 3 min read NVDA AEHR Aehr Test Systems Q4 2026 Earnings Call Summary - Moby Strategic Market Diversification and AI Momentum Our analysts just identified a stock with the potential to be the next Nvidia. Tell us how you invest and we'll show you why it's our #1 pick. Tap here.

Successfully transitioned from 95% electric vehicle silicon carbide revenue to 95% non-EV revenue, primarily driven by AI processors and data center infrastructure. AI processor wafer-level burn-in emerged as the fastest-growing segment, representing approximately 71% of total annual revenue due to increased capacity needs from lead customers. Achieved a critical technical milestone by completing benchmark testing for a top-tier AI processor supplier, demonstrating results superior to traditional package-level testing.

Silicon photonics demand is accelerating as AI data center architectures shift toward optical I/O and high-speed interconnects, shortening the sales cycle for new networking customers. Maintained a competitive advantage in power semiconductors by winning a silicon carbide customer in Taiwan over a Chinese competitor based on technical superiority and automotive reputation. Strategic focus on wafer-level burn-in is driven by its ability to manage individual die temperatures and prevent yield loss in high-value multi-chip packages.

Fiscal 2027 Growth Outlook and Capacity Expansion Projected fiscal 2027 revenue of $130 million to $150 million, representing roughly 2. 6x to 3x growth over fiscal 2026, supported by a record $100. 6 million effective backlog.

Guidance assumes non-GAAP pretax profitability between 18% and 22%, with potential upside if additional orders from the new AI benchmark customer materialize. Manufacturing capacity is being aggressively scaled through a contract manufacturer in Southeast Asia capable of producing 20 Sonoma systems per month to meet hyperscale demand. Strategic roadmap includes expanding into the memory market, specifically targeting high-bandwidth memory (HBM) and flash, with potential orders in fiscal 2027 and ramps in fiscal 2028.

Management is prioritizing a capital-light model by using contract manufacturers to handle lumpy demand cycles without building excessive fixed permanent infrastructure. Operational Risks and Legal Developments Ongoing patent infringement litigation against Semi/Nexus Test in China remains a focus, with the Beijing patent office recently upholding two of Aehr's critical patents. Supply chain constraints for high-power components used in Sonoma systems have led to 40% price increases from certain vendors who also supply major AI chipmakers.

The company significantly strengthened its balance sheet by raising approximately $100 million through an ATM program to support larger customer opportunities. Expansion of the Hsinchu, Taiwan office is underway to deepen engagement with the regional semiconductor ecosystem and support upcoming production ramps. Story Continues Q&A Session Highlights Revenue mix and product assumptions for fiscal 2027 guidance One stock.

Nvidia-level potential. 30M+ investors trust Moby to find it first. Get the pick.

Tap here. Management expects the mix to remain similar to the prior year, with AI representing roughly 70%, silicon photonics 15-20%, and power semiconductors making up the remainder. The current $130 million to $150 million guidance does not include any revenue from the memory market or the recently completed AI benchmark customer.

Development progress and timing for the memory market entry Aehr has begun spending on R&D for memory, including architectural schematics and hiring, but requires a committed customer agreement before full deployment. New HBM standards with embedded 'Built-In Self Test' (BIST) capabilities may allow Aehr to enter the DRAM market sooner than originally anticipated. Total revenue capacity and manufacturing pinch points The Fremont facility has been upgraded to support significantly higher throughput, potentially handling $100 million in wafer-level systems per month if fully staffed.

The company is using a 'hybrid' capacity model where Sonoma systems are outsourced to Asia, freeing up the Fremont floor for complex wafer-level FOX systems. Transition from package-level to wafer-level burn-in for AI A lead hyperscale customer is evaluating wafer-level burn-in for their third-generation device to avoid the massive yield loss associated with failing a single die in a multi-chip package. Wafer-level systems carry higher price points and better margins than package-level systems due to the significant R&D and IP involved.