Cohu (COHU) Q1 2026 Earnings Call Transcript

Cohu reported strong Q1 results with orders up 57% year over year, revenue of $125.1 million, and gross margin of 46.5%, while raising 2026 HPC revenue guidance to $80 million-$100 million and full-year revenue growth to 20%-25%. The computing pipeline is now $750 million, with $100 million in near-term qualified opportunity and another $200 million in qualification, led by Eclipse thermal handlers and HBM inspection. The main offset is margin pressure from higher Eclipse ramp costs, which should keep gross margin in the mid-40% range and OpEx in the low-$50 million range for the rest of 2026.

Analysis

COHU is not just participating in the AI capex cycle; it is moving from a cyclical tester into a semi-embedded infrastructure layer for power-dense compute. The key second-order effect is that qualification success in Eclipse and Neon should widen the moat: once handlers are standardized into customer process flows, the switching cost is less about unit price and more about yield protection, which supports future software attach and post-warranty revenue. That creates a multi-year monetization stack: systems today, consumables/services next, and software later.

The near-term tradeoff is margin compression for share gain. Management is effectively choosing to harvest a larger addressable pool now even though supply-chain ramps and mix shift pressure gross margin into the mid-40s; that is usually the right call if it converts into installed-base lock-in, but it also means the stock may need proof of conversion before rerating further. The most important variable for the next 2 quarters is not demand, it is throughput: if the 14-week production cadence slips, 2026 upside can easily get pushed into 2027 and the market will reprice the name as a delayed story rather than a clean acceleration.

The contrarian miss is that consensus may be underestimating how broad the AI test bottleneck is beyond GPU headlines. If compute demand is spreading across CPUs, custom ASICs, network processors, and power devices, COHU’s opportunity set becomes less dependent on a single platform cycle and more on the structural buildout of AI datacenter power and thermal management. That said, the pipeline is still highly qualified rather than converted, so the stock is vulnerable to any signal that qualification wins do not translate into backlog fast enough to offset the second-half margin drag.

Key Decisions for Investors

• Long COHU on a 3-6 month horizon; use pullbacks toward post-earnings support to build a position. Base case is multiple expansion as 2026 revenue visibility improves, with upside if Q2/Q3 confirm conversion from qualification to backlog.
• Buy COHU Jan-2027 calls or call spreads to express the multi-quarter qualification thesis while limiting downside to the margin-compression risk. Best risk/reward if the market is underpricing 2027 revenue carry from the current pipeline.
• Pair trade: long COHU / short a lower-quality semiconductor equipment name with less recurring content. The trade favors names with installed-base monetization and software attach, which should outperform if the market begins to reward durability over near-term margin purity.
• Hedge with a short-dated put spread against COHU if entering before the next commentary point. The main risk is a slip in shipping cadence or a weaker-than-expected Q3 conversion rate, which would likely compress the multiple quickly.
• For broader exposure, add a small basket long in AI infrastructure enablers where COHU is the most asymmetric thermal/power bottleneck play; keep sizing modest because the upside depends on execution, not just end-market demand.