Can AEHR's Diversification Strengthen Its Growth Outlook?

Aehr Test Systems has shifted its revenue mix sharply toward AI, with AI accelerators, CPUs and network processors contributing about 71% of fiscal 2026 revenue, versus more than 95% of its business tied to EV silicon carbide two years earlier. The company secured a $22 million follow-on AI wafer-level burn-in order in August and roughly $8 million in SiC orders in July, supporting its $100.6 million backlog and fiscal 2027 revenue outlook of $130 million-$150 million. Fiscal 2027 EPS consensus is $0.74 versus $0.03 a year earlier, though AEHR trades at a rich 19.05x forward sales multiple and faces strong competition from Teradyne and FormFactor.

Analysis

AEHR's investment case is now a concentration-risk trade disguised as diversification: AI-related demand may broaden end markets, but it also ties revenue recognition to a small number of customer qualification and production-ramp decisions. Wafer-level burn-in is particularly lumpy because systems orders can precede utilization and repeat purchases by several quarters. The relevant near-term KPI is not backlog alone, but backlog conversion, customer concentration, gross-margin progression, and whether new orders represent capacity expansion rather than replacement or pilot tooling.

The competitive asymmetry favors TER and FORM as more liquid, diversified ways to express AI test intensity. AEHR's niche can produce superior upside if wafer-level burn-in becomes a standard reliability step for increasingly power-dense AI processors and optical interconnects; however, large ATE vendors can bundle adjacent test workflows, while probe-card suppliers benefit regardless of whether customers choose AEHR's architecture. A delayed HBM/logic test cycle would likely hurt AEHR's multiple disproportionately given its premium revenue valuation and recent momentum.

Over the next 1-3 months, follow-on production orders and fiscal-year guidance reaffirmation can sustain the narrative, but expectations leave limited tolerance for a single order delay. Over 6-18 months, the key structural question is whether silicon photonics and GaN become repeatable revenue verticals that reduce cyclicality, versus a collection of low-volume evaluations. Contrarian view: the market may be capitalizing peak AI equipment order rates before evidence of recurring service revenue or a broader customer base; positive AI headlines alone are insufficient to justify further multiple expansion.

Key Decisions for Investors

• Do not chase AEHR after the momentum move; place on a 1-3 month watch for the next earnings release and initiate only if management reaffirms revenue outlook while backlog conversion and gross margin improve sequentially. A guidance cut or material pushout in AI production orders falsifies the long thesis.
• Express AI semiconductor test exposure through a long FORM / short AEHR pair over 3-6 months: FORM offers broader HBM, foundry and logic test participation, while AEHR carries greater customer-concentration and valuation-decompression risk. Size modestly because a large AEHR follow-on order can create sharp short-term relative upside.
• For a higher-beta AEHR long, use defined-risk call spreads dated 6-9 months rather than common equity, contingent on confirmation that additional AI or photonics orders are production deployments. Target upside requires sustained order momentum; maximum loss should be limited to premium if qualification programs fail to convert.
• Maintain TER as the lower-volatility long proxy for AI test spending through the next two reporting cycles. Reassess if Semiconductor Test orders or HBM-related commentary weaken, which would indicate a broader capex digestion rather than an AEHR-specific execution issue.