Cohu Inc (COHU) Stock Analysis and Metrics

Cohu

NASDAQ: COHU

$52.09 ▲ +1.12 (+2.20%)

At close: Jul 8, 2026 · 3:59 PM UTC

Financial Ratios

Market Cap3.25 Bn

P/E-36.72

P/S6.75

Div. Yield0.00

ROIC (Qtr)0.00

Total Debt (Qtr)286.20 Mn

Revenue Growth (1y) (Qtr)29.26

Add ratio to table…

About

Cohu, Inc. is a global supplier of equipment and services that optimize semiconductor manufacturing yield and productivity. The company provides test automation systems that move devices and control temperature during test, inspection and metrology tools for molded leaded and leadless devices and bare dies, semiconductor automated test equipment for radio frequency digital mixed signal and power management applications, interface products such as test contactors and power probe cards, software analytics including an AI driven digital twin platform, and services covering device applications spares board repair and service agreements. Its products enable connectivity autonomous driving high performance computing for AI applications advanced medical equipment and robotic automation. Cohu maintains an active installed base exceeding 25,000 systems serving over 280 high volume manufacturing facilities across 108 customers in 31 countries.

Cohu generates revenue from the sale of semiconductor test and inspection systems including kits and from recurring revenue streams such as interface products spares upgrades configuration tooling software analytics and services. The recurring revenue model benefits from a large installed base that drives steady demand for consumables and support services. In fiscal 2025 systems represented 40% of net sales while recurring revenues accounted for 60% of net sales. The company’s research and development expense was $92.2 million in fiscal 2025 $84.8 million in fiscal 2024 and $88.6 million in fiscal 2023 reflecting ongoing investment in next generation technologies.

The company operates through a single reportable segment named Semiconductor Test and Inspection Equipment. This segment encompasses all of Cohu’s product lines and service offerings.

• Semiconductor Test and Inspection Equipment: This segment provides test automation systems that automate device movement and temperature control, inspection and metrology equipment for molded leaded and leadless devices and bare dies, semiconductor automated test equipment for radio frequency digital mixed signal and power management applications, interface products such as test contactors and power probe cards, software analytics including an AI driven digital twin platform that delivers real time online performance monitoring prescriptive maintenance and machine learning process control, and services covering device applications spares board repair and service agreements supported by a global field service team.

Cohu holds a strong position in the competitive semiconductor test automation and inspection market alongside major players such as Advantest Teradyne Hon Precision and KLA. The company differentiates itself through a broad and integrated product portfolio that spans test automation inspection interface solutions thermal subsystems and software analytics. Significant investments in research and development focus on high growth areas including high performance computing automotive applications and high bandwidth memory. Proprietary technologies in active thermal control vision inspection and machine learning based process control give Cohu an edge in testing complex devices. A resilient recurring revenue model driven by services consumables and software analytics provides stable cash flow. Strategic partnerships targeted acquisitions and a disciplined approach to capital allocation enable expansion into new markets and technologies. A large global field service organization supports customers worldwide and enhances customer loyalty.

Cohu serves integrated device manufacturers outsourced semiconductor assembly and test providers and fabless semiconductor companies. Among its customers STMicroelectronics accounted for 10% or more of consolidated net sales in recent years making it a significant customer. The company’s customer base spans the global semiconductor ecosystem with presence in the Americas Europe and Asia Pacific. Customers rely on Cohu’s products to achieve high yield improved productivity and lower cost of test in markets such as computing and networking automotive industrial mobile and consumer electronics.

Investment Thesis

▲ Bull case

Cohu's strategic positioning in AI infrastructure power management represents an underappreciated growth catalyst beyond the commonly discussed compute and memory segments. The recent $5 million Diamond X platform order from a leading semiconductor manufacturer for GaN power device testing reveals deep penetration into the AI data center power architecture value chain—a critical enabler as server power density increases with GaN adoption. This addresses a bottleneck where power conversion inefficiencies directly limit AI system performance, yet receives less market attention than GPU/CPU testing. Management's emphasis on Diamond X supporting the full GaN lifecycle from development to high-volume production indicates recurring revenue potential through consumables and software as customers scale deployment. Given that AI infrastructure spends disproportionately on power delivery relative to compute (often 30-40% of total power budget), and Cohu's proprietary thermal and precision test capabilities are uniquely suited for GaN characterization, this segment could materially exceed current guidance. The company's $750 million computing segment opportunity pipeline likely undercounts power-related opportunities, creating asymmetric upside if management successfully converts these engagements into revenue streams with higher gross margins than compute-focused offerings.
The software subscription model demonstrates significant latent value that current financials fail to capture, with management citing a $20 million system order generating $330,000 annually in software—translating to ~$5 million lifetime revenue per major win. While Q1 ARR stood at only $1.2 million with a low 1.3% attachment rate, this reflects early-stage adoption rather than limited demand. The transition from pilot to production environments for analytics software, combined with Cohu's focus on prescriptive maintenance and yield optimization, creates a flywheel where each system sale increases switching costs through operational stickiness. As noted in the transcript, software provides high-margin recurring revenue less susceptible to CapEx cycles—a critical advantage in the cyclical semiconductor equipment industry. With R&D investments accelerating and production capacity expanding to support HPC opportunities, the software attachment rate is poised for non-linear growth as customers prioritize quality and scalability in AI data center applications. This recurring revenue stream could significantly improve earnings quality and reduce revenue volatility over time, yet the market appears to be valuing Cohu primarily on its hardware cycle.
Cohu's manufacturing ramp and supply chain investments position it to capture delayed qualification benefits that could accelerate 2027 growth beyond current expectations. Management explicitly stated that while they aim to convert some of the ~$100 million qualified opportunity in 2026, the larger percentage will likely convert in 2027 due to six-month qualification timelines plus production ramp. The company is proactively addressing this by hiring resources in Malaysia, re-laying facilities for increased floor space, and working with suppliers to resolve choke points—particularly for thermal heads. This preparation suggests Cohu could convert more of its $150-$200 million 2027 qualification bucket into revenue sooner than anticipated if supply chain bottlenecks ease faster than expected. Furthermore, the 14-week cycle time for thermal handlers, while seemingly long, is manageable given the backlog-driven nature of lead times, and ongoing operational improvements could compress this timeline. With test utilization already at 78% and industrial markets showing pickup (79% utilization at Q1 end), the inflection point for broader semiconductor recovery may be nearer than modeled, allowing Cohu to benefit from both its HPC specialization and recovering end-markets simultaneously.

Cohu's strategic positioning in AI infrastructure power management represents an underappreciated growth catalyst beyond the commonly discussed compute and memory segments. The recent $5 million Diamond X platform order from a leading semiconductor manufacturer for GaN power device testing reveals deep penetration into the AI data center power architecture value chain—a critical enabler as server power density increases with GaN adoption. This addresses a bottleneck where power conversion inefficiencies directly limit AI system performance, yet receives less market attention than GPU/CPU testing. Management's emphasis on Diamond X supporting the full GaN lifecycle from development to high-volume production indicates recurring revenue potential through consumables and software as customers scale deployment. Given that AI infrastructure spends disproportionately on power delivery relative to compute (often 30-40% of total power budget), and Cohu's proprietary thermal and precision test capabilities are uniquely suited for GaN characterization, this segment could materially exceed current guidance. The company's $750 million computing segment opportunity pipeline likely undercounts power-related opportunities, creating asymmetric upside if management successfully converts these engagements into revenue streams with higher gross margins than compute-focused offerings.
The software subscription model demonstrates significant latent value that current financials fail to capture, with management citing a $20 million system order generating $330,000 annually in software—translating to ~$5 million lifetime revenue per major win. While Q1 ARR stood at only $1.2 million with a low 1.3% attachment rate, this reflects early-stage adoption rather than limited demand. The transition from pilot to production environments for analytics software, combined with Cohu's focus on prescriptive maintenance and yield optimization, creates a flywheel where each system sale increases switching costs through operational stickiness. As noted in the transcript, software provides high-margin recurring revenue less susceptible to CapEx cycles—a critical advantage in the cyclical semiconductor equipment industry. With R&D investments accelerating and production capacity expanding to support HPC opportunities, the software attachment rate is poised for non-linear growth as customers prioritize quality and scalability in AI data center applications. This recurring revenue stream could significantly improve earnings quality and reduce revenue volatility over time, yet the market appears to be valuing Cohu primarily on its hardware cycle.
Cohu's manufacturing ramp and supply chain investments position it to capture delayed qualification benefits that could accelerate 2027 growth beyond current expectations. Management explicitly stated that while they aim to convert some of the ~$100 million qualified opportunity in 2026, the larger percentage will likely convert in 2027 due to six-month qualification timelines plus production ramp. The company is proactively addressing this by hiring resources in Malaysia, re-laying facilities for increased floor space, and working with suppliers to resolve choke points—particularly for thermal heads. This preparation suggests Cohu could convert more of its $150-$200 million 2027 qualification bucket into revenue sooner than anticipated if supply chain bottlenecks ease faster than expected. Furthermore, the 14-week cycle time for thermal handlers, while seemingly long, is manageable given the backlog-driven nature of lead times, and ongoing operational improvements could compress this timeline. With test utilization already at 78% and industrial markets showing pickup (79% utilization at Q1 end), the inflection point for broader semiconductor recovery may be nearer than modeled, allowing Cohu to benefit from both its HPC specialization and recovering end-markets simultaneously.

▼ Bear case

Cohu's gross margin trajectory faces significant headwinds from mix shift and operational inefficiencies that management may be underestimating. While Q1 gross margin exceeded guidance at 46.5% due to favorable recurring revenue mix, CFO Jeffrey Jones explicitly warned that increasing systems revenue—particularly from the lower-margin Eclipse platform ramp—will pressure margins toward the mid-40% range for FY26. This is compounded by higher initial costs for Eclipse supply chain and production, which Jeff stated will "carry through probably through this year," alongside 20 basis points of headwinds from elevated memory IC and energy/freight costs. The market may be overlooking that as HPC systems revenue grows faster than recurring (currently 60% of revenue), the blended gross margin decline could accelerate beyond management's guidance, especially if Eclipse-related operational hiccups persist. Furthermore, the company's reliance on converting a $150-$200 million qualification pipeline into 2027 revenue assumes smooth execution, yet Luis Müller acknowledged uncertainties around lead times, backlog impact, and whether qualifications completed in early Q3 can translate to 2026 revenue recognition given accounting rules. Any delay in this conversion would directly impact 2027 growth prospects, yet the stock may be pricing in near-perfect execution of this transition.
Concentration risk in Cohu's high-growth opportunities presents a material threat that was not adequately addressed during Q&A, despite multiple references to specific customer wins. Luis Müller highlighted strategic Eclipse wins with two U.S.-based fabless customers representing ~$100 million incremental revenue over three years each, while inspection growth relies heavily on repeat orders from a U.S.-headquartered and a Korean customer for NEON systems. This creates significant customer concentration in the very segments driving upward revisions to guidance. The transcript reveals that no single customer exceeded 10% of Q1 sales, but this masks potential future concentration as large wins materialize—particularly since the $750 million computing segment opportunity pipeline is based on a "very targeted list" of just 15 customers. If key accounts delay qualifications, shift to competitors, or experience internal budget cuts (e.g., due to AI infrastructure overbuild concerns), Cohu's revenue outlook could deteriorate rapidly. Management's focus on expanding engagement with existing strategic accounts, rather than diversifying the customer base, increases vulnerability to idiosyncratic risks at these flagship clients, yet the market appears to be treating the HPC opportunity as broadly diversified across the semiconductor landscape.
The semiconductor equipment industry's inherent cyclicality and Cohu's rising operating expenses create a precarious financial profile that could worsen if market conditions deteriorate faster than anticipated. Despite Q1 non-GAAP income of only $0.6 million, OpEx increased to $55 million (above guidance) to support HPC opportunities, with Jeff Jones stating OpEx will remain in the "low-$50 million range" through 2026—a level that requires sustained revenue growth to maintain profitability. The company carries $305 million in total debt ($288 million from convertible notes), and while net interest income provided ~$2.1 million benefit in Q1, this is sensitive to interest rate fluctuations and foreign currency impacts. Crucially, Cohu's guidance assumes Q2 revenue will increase 15% sequentially to ~$144 million, with flat revenue through the balance of the year needed to hit the 20%-25% FY26 growth target. If automotive/industrial recovery stalls or HPC qualification timelines slip further—as hinted by Luis Müller's uncertainty about Q4 versus early 2027 revenue recognition—revenue could flatline or decline, triggering operating leverage losses. Given that R&D and field support investments are being scaled to capture opportunities (as noted in the OpEx discussion), any revenue shortfall would not be quickly offset by cost cuts, potentially leading to prolonged margin pressure and cash burn that the market has not fully priced in despite the company's strong balance sheet ($488.7 million cash/investments).

Cohu's gross margin trajectory faces significant headwinds from mix shift and operational inefficiencies that management may be underestimating. While Q1 gross margin exceeded guidance at 46.5% due to favorable recurring revenue mix, CFO Jeffrey Jones explicitly warned that increasing systems revenue—particularly from the lower-margin Eclipse platform ramp—will pressure margins toward the mid-40% range for FY26. This is compounded by higher initial costs for Eclipse supply chain and production, which Jeff stated will "carry through probably through this year," alongside 20 basis points of headwinds from elevated memory IC and energy/freight costs. The market may be overlooking that as HPC systems revenue grows faster than recurring (currently 60% of revenue), the blended gross margin decline could accelerate beyond management's guidance, especially if Eclipse-related operational hiccups persist. Furthermore, the company's reliance on converting a $150-$200 million qualification pipeline into 2027 revenue assumes smooth execution, yet Luis Müller acknowledged uncertainties around lead times, backlog impact, and whether qualifications completed in early Q3 can translate to 2026 revenue recognition given accounting rules. Any delay in this conversion would directly impact 2027 growth prospects, yet the stock may be pricing in near-perfect execution of this transition.
Concentration risk in Cohu's high-growth opportunities presents a material threat that was not adequately addressed during Q&A, despite multiple references to specific customer wins. Luis Müller highlighted strategic Eclipse wins with two U.S.-based fabless customers representing ~$100 million incremental revenue over three years each, while inspection growth relies heavily on repeat orders from a U.S.-headquartered and a Korean customer for NEON systems. This creates significant customer concentration in the very segments driving upward revisions to guidance. The transcript reveals that no single customer exceeded 10% of Q1 sales, but this masks potential future concentration as large wins materialize—particularly since the $750 million computing segment opportunity pipeline is based on a "very targeted list" of just 15 customers. If key accounts delay qualifications, shift to competitors, or experience internal budget cuts (e.g., due to AI infrastructure overbuild concerns), Cohu's revenue outlook could deteriorate rapidly. Management's focus on expanding engagement with existing strategic accounts, rather than diversifying the customer base, increases vulnerability to idiosyncratic risks at these flagship clients, yet the market appears to be treating the HPC opportunity as broadly diversified across the semiconductor landscape.
The semiconductor equipment industry's inherent cyclicality and Cohu's rising operating expenses create a precarious financial profile that could worsen if market conditions deteriorate faster than anticipated. Despite Q1 non-GAAP income of only $0.6 million, OpEx increased to $55 million (above guidance) to support HPC opportunities, with Jeff Jones stating OpEx will remain in the "low-$50 million range" through 2026—a level that requires sustained revenue growth to maintain profitability. The company carries $305 million in total debt ($288 million from convertible notes), and while net interest income provided ~$2.1 million benefit in Q1, this is sensitive to interest rate fluctuations and foreign currency impacts. Crucially, Cohu's guidance assumes Q2 revenue will increase 15% sequentially to ~$144 million, with flat revenue through the balance of the year needed to hit the 20%-25% FY26 growth target. If automotive/industrial recovery stalls or HPC qualification timelines slip further—as hinted by Luis Müller's uncertainty about Q4 versus early 2027 revenue recognition—revenue could flatline or decline, triggering operating leverage losses. Given that R&D and field support investments are being scaled to capture opportunities (as noted in the OpEx discussion), any revenue shortfall would not be quickly offset by cost cuts, potentially leading to prolonged margin pressure and cash burn that the market has not fully priced in despite the company's strong balance sheet ($488.7 million cash/investments).

Segments Breakdown of Revenue (2025)

Peer Comparison

Companies in the Semiconductor Equipment & Materials

S.No.	Ticker	Company	Market Cap	P/E	P/S	Total Debt (Qtr)
1	AMAT	Applied Materials Inc /De	516.82 Bn	60.75	17.81	6.46 Bn
2	LRCX	Lam Research Corp	488.97 Bn	72.89	22.55	3.73 Bn
3	KLAC	Kla Corp	348.47 Bn	74.61	26.61	-
4	TER	Teradyne, Inc	66.84 Bn	70.06	17.65	-
5	Q	Qnity Electronics, Inc.	32.19 Bn	47.61	6.57	4.02 Bn
6	ENTG	Entegris Inc	25.16 Bn	94.72	7.78	3.65 Bn
7	AMKR	Amkor Technology, Inc.	19.80 Bn	45.18	2.80	1.41 Bn
8	FORM	Formfactor Inc	11.45 Bn	166.30	13.63	0.01 Bn

Research Tools

Developers

About

Investment Thesis

Geographical Breakdown of Revenue (2025)

Segments Breakdown of Revenue (2025)

Peer Comparison

Research Tools

Developers

About

Investment Thesis

Geographical Breakdown of Revenue (2025)

Segments Breakdown of Revenue (2025)

Peer Comparison

Create an account

Sign In

Reset Password