Few small-cap equipment makers have opened a fiscal year with a bigger gap between what management says and what the market seems willing to believe. Last quarter Aehr Test Systems laid out fiscal 2027 revenue of $130 million to $150 million, roughly two and a half to three times the $50 million it booked in fiscal 2026. Analysts are still penciling in about $86.6 million for the year, far below even the bottom of that range, and the stock has since lost 43.7% while the S&P 500 edged higher. The first quarter of the new year is the first real chance to show whether that guide was a forecast or a hope.

For the August quarter, consensus calls for non-GAAP earnings of $0.11 a share on revenue of $27.5 million. That would be a 150% jump from $11.0 million a year ago, when the company lost $0.05 a share, and a solid step up from the $18.8 million reported last quarter. The whisper number sits at $0.15, a meaningful premium to consensus for a company of this size, suggesting the more engaged crowd expects volume and mix to lift profitability faster than the published models assume. Even so, $27.5 million would be a modest start for a $130 million year. Hitting the low end would require the remaining three quarters to average more than $34 million, so the tone of commentary on second-half shipment timing may matter as much as the headline figures.

The raw material for a ramp is there. Aehr exited fiscal 2026 with a record $80.6 million backlog, closer to $100 million including orders received shortly after the quarter, driven by a $41 million Sonoma order from its lead hyperscaler and record bookings above $60 million. What validates the story now is conversion. Revenue landing at or above consensus would show that backlog is moving through the factory on schedule, and fresh bookings that keep backlog roughly intact despite heavier shipments would argue that demand is broadening rather than concentrated in one giant purchase order. A sharp drop in backlog without replacement orders would revive concerns about lumpiness, especially with three customers each accounting for more than 10% of revenue last quarter.

Margins are the second test. Non-GAAP gross margin climbed to 45% last quarter from below 30% two quarters earlier, helped by volume and utilization, and management framed fiscal 2027 pre-tax margins at 18% to 22%. Holding gross margin in the mid-40s would support that framework. Slippage would point to the headwinds management flagged, including power supply costs up around 40% amid competition for AI components and rising legal bills from patent litigation in China.

Then there are the catalysts outside the guide. The top-tier AI processor benchmark, which had stumbled for two quarters, finished with results management described as better than package-level testing, and the customer is moving a high-volume device toward pilot production in Taiwan while evaluating a second. Any order from that customer would be pure upside. Progress from the new silicon photonics networking customer, further silicon carbide orders after roughly $8 million arrived in a single month, and early traction in gallium nitride wafer-level burn-in would all reinforce the diversification argument. Memory remains excluded from the outlook, so delays there should not move the needle.

Sentiment has warmed to 30.6% bullish from 12.5% a quarter ago, yet the shares trade near $55, barely above the post-earnings low of $54.22, well below the $73.30 200-day average and a long way from the $147.40 peak reached after the last report. That combination suggests investors like the narrative but have stopped paying for it. The central question is simple: will management reaffirm or raise the $130 million to $150 million range with backlog and bookings that make it credible? If it does, the distance between consensus and guidance could narrow quickly. If the range wobbles, the market's skepticism will look justified.