1 Under-the-Radar AI Semiconductor Stock to Buy Hand Over Fist, According to Wall Street

Cohu is forecast to accelerate revenue growth to $558.5 million in 2026, up 23% year over year, following 13% growth in 2025, while its high-performance computing pipeline totals $750 million. All seven Wall Street Journal-tracked analysts rate the stock a buy, with an average target of $57.43 implying 24% upside and a Street-high target of $65 implying 40% upside. The article argues AI-related demand for Cohu's semiconductor testing and inspection equipment could drive further earnings growth into 2027 and beyond.

Analysis

The important second-order read-through is that this is not an AI-chip beta trade; it is a capacity-validation trade. When hyperscalers and memory vendors move from design wins to qualification, equipment names like this become the hidden leverage on the buildout, and the revenue inflection tends to arrive in steps rather than a smooth line. That means the market may be underestimating how quickly backlog can convert once qualification clears, but also overestimating near-term linearity if a few large accounts slip by a quarter or two.

The broader winner set is the upstream semiconductor equipment ecosystem that benefits from rising test complexity and tighter defect tolerances, especially in HBM and advanced packaging. If the testing/inspection layer is seeing 60%+ order growth, that usually signals the industry is moving from proof-of-concept AI spend into industrialized deployment, which should also support consumables, software-driven metrology, and adjacent automation vendors. The less obvious loser is any chipmaker that is still relying on older yield assumptions; as defect detection improves, weak processes get exposed faster, potentially raising scrappage and near-term gross margin volatility before yields improve.

The main risk is not demand disappearing, but timing compression: the stock can de-rate if the market has already capitalized 2027-28 earnings before those numbers are visible. At the current setup, the valuation gap versus the larger AI semis is less about cheap vs expensive and more about confidence in multi-year conversion of pipeline into booked revenue. If qualification cycles extend, the multiple can compress even while the secular thesis remains intact.

Consensus appears to be missing that the best way to express this theme may be through a basket or pair, not a standalone long. The catalyst path is likely 6-18 months, with near-term upside tied to order conversion and next-year guidance, while the bigger upside depends on sustained HBM/AI test intensity into 2027. The move looks directionally right but tactically crowded after a major run, so the better edge is on pullbacks or through relative-value structures.

Key Decisions for Investors

• Buy COHU on a pullback of 8-12% from recent highs with a 6-12 month horizon; target a re-rating as pipeline conversion becomes visible, but size modestly because the stock has already repriced a meaningful portion of the good news.
• Pair trade: long COHU / short a basket of mature semi-cap equipment names with lower AI exposure over 3-6 months; thesis is that AI test/inspection mix should outperform general fab equipment as memory complexity rises.
• Add NVDA and MU on weakness rather than chasing COHU outright; if COHU’s order strength is real, it is a confirming signal for sustained AI capex and HBM demand over the next 2-4 quarters.
• Use call spreads in COHU for a 6-9 month view instead of common stock; implied upside is meaningful, but the already-expanded multiple makes upside more path-dependent than usual.