The semiconductor map: who really manufactures, who inspects, and who only supplies components

Why The Value Chain Needs To Be Mapped

The word "semiconductors" creates a tempting shortcut. Investors see a company that supplies equipment to fabs, a company that manufactures boards, or a company that sells electronic components, and all of them are quickly placed under the same heading. That is not how the economics work. A chip manufacturer, a test-equipment company, a substrate supplier, a component distributor and a small assembly subcontractor do not move with the same cycle and do not deserve the same analytical treatment.

The local market has no giant chip index, but it does have a surprisingly broad map. It includes Tower, which is the only true local manufacturing anchor; Nova, Camtek, Qualitau and CI Systems, which sell metrology, inspection, reliability and control equipment; Priortech and PCB Technologies, which connect to packaging, substrates and miniaturization; Telsys, which is exposed through components and modules; Arbe, which designs radar chips but is still before meaningful commercial production; and Ashot, Imco and DSIT, which supply industrial assemblies or electronics with only partial semiconductor exposure.

That distinction matters. When demand for AI infrastructure, HBM, advanced packaging and data centers rises, the benefit does not reach every company in the same way. Some companies sell directly into the investment cycle. Others feel it through backlog, component availability, one large customer, or a small segment that sits inside a broader defense or industrial company. The useful question is therefore not "who is connected to chips?", but where each company sits in the chain and how much of its financial profile is actually affected.

The Value-Chain Map

Tower: The Local Manufacturing Anchor

Tower is the clearest semiconductor company on the local exchange because it sits at the manufacturing layer. It operates as an analog foundry and manufactures integrated circuits for customers in areas such as Silicon Photonics, SiGe, BiCMOS, RF CMOS, image sensors, displays and power management. Unlike companies that sell equipment to the industry, Tower is inside the production chain itself.

In 2025, Tower's revenue rose to $1.566 billion, compared with $1.436 billion in 2024, an increase of about 9%. Gross profit was $363.9 million, operating profit was $194.2 million, and annual profit attributable to shareholders was $220.5 million. The company attributes the increase mainly to higher wafer shipments, including a full-year contribution from the Agrate facility in Italy.

The core metric for Tower is not only revenue growth. In manufacturing, the key questions are utilization, product mix, depreciation, manufacturing costs and how much revenue converts into gross and operating profit. Tower therefore gives local investors the most direct exposure to semiconductor manufacturing, but it is also a mature industrial company. It should be read through cycle, capacity, customer mix and execution, not as an option on a new technology.

Metrology And Inspection: The Deepest Local Group

The most substantial local cluster is not chip manufacturing, but metrology, inspection, process control and reliability testing. That is not a side issue. As chips, packages and production processes become more complex, mistakes become more expensive. More steps require more measurement, more inspection and more control before defective production turns into a costly loss.

Nova focuses on metrology and process control at global scale

Nova is one of the strongest local examples of pure semiconductor-equipment exposure. The company develops metrology and process-control systems for chip manufacturing, and essentially all of its revenue comes from this industry. It does not manufacture chips, but it sells systems that allow manufacturers to monitor the production process, measure wafers, identify deviations and improve yield.

In 2025, Nova reported revenue of $880.6 million, up 31% from 2024. Product revenue was $705.6 million and service revenue was $175.0 million. Gross margin was 57%, while both operating profit and net profit were 29% of revenue. The company also entered 2026 with a very strong balance sheet: cash, deposits and marketable securities of $1.646 billion, compared with $820 million at the end of 2024.

The risk is that Nova is tied to semiconductor capital spending. Customers may change or delay planned shipments, and backlog is not a full guarantee of near-term revenue. But the 2025 figures show a company with scale, profitability and financial flexibility. On the local map, Nova belongs in the core group.

Camtek offers the sharpest exposure to AI, HBM and advanced packaging

Camtek is also a semiconductor-equipment company, but with a different angle. It sells inspection and metrology systems, with strong exposure to advanced packaging. In the current cycle, that is especially important because more performance comes not only from smaller chips, but also from combining chips, memory and substrates in more complex packages that require more inspection.

In 2025, Camtek reported revenue of $496.1 million, compared with $429.2 million in 2024 and $315.4 million in 2023. Operating profit reached $128.2 million and annual profit was $118.5 million. The company's most important disclosure is that about 50% of revenue came from products supporting AI applications, mainly HBM and advanced packaging.

This makes Camtek one of the most direct local beneficiaries of the AI infrastructure cycle. But it also creates concentration risk. If AI, HBM and chiplet-related demand keeps growing, Camtek remains in the right place in the value chain. If the cycle slows, this same exposure can work against it. Geographic exposure also matters: 49% of 2025 revenue came from China. The company therefore faces not only a technology-cycle test, but also a geographic and regulatory one.

At Qualitau, each test-equipment order can move the numbers

Qualitau is much smaller than Nova and Camtek, but it sits in an important niche: reliability testing for semiconductor wafers and packages. As chip architectures and advanced packaging become more complex, reliability testing becomes a prerequisite for production, not an optional add-on.

In 2025, Qualitau reported revenue of $61.3 million and gross profit of $43.6 million, a gross margin of 71.2%. The more important number is backlog: $42.3 million at the end of 2025, rising to $62.3 million at the end of March 2026. The company reports orders from customers in the United States and Asia, with most of the March 2026 backlog expected to be delivered in the fourth quarter of 2026 and the first quarter of 2027.

Because Qualitau is small, a few million dollars of new orders can matter. The same is true in the other direction: delivery delays or order timing can weigh on a full year. The company is focused and profitable, but the investment case is more volatile and depends heavily on order flow, delivery timing and customer decisions.

At CI Systems, microelectronics is still part of a broader company

CI Systems is not a classic semiconductor name, but the microelectronics activity grew enough in 2025 to justify inclusion on the map. The company develops and markets electro-optical test and metrology equipment for the semiconductor industry, including the WETALYZER system for measuring liquids used in manufacturing processes.

In 2025, consolidated revenue rose 44.3% to $48.5 million. Gross profit was $17.3 million and annual profit was $2.0 million. The microelectronics activity generated $10.6 million of revenue, representing 22% of sales, compared with $6.5 million in 2024. The backlog in this activity was stable at $2.75 million at the end of 2025 and at the end of March 2026.

The semiconductor exposure is visible, but it is still part of a company where the military-scientific business accounts for 77% of sales. CI Systems should not be read like Nova or Camtek. It should be read as an industrial company where microelectronics can improve mix and profitability if it continues to grow, but does not yet define the whole story.

Packaging, Substrates And Boards Are Becoming Critical

As the semiconductor world advances, not every improvement comes from shrinking transistors. More value is created in the way chips, memories, substrates and systems are connected, how heat is dissipated, how signals move and how density is preserved. Substrates, advanced packaging and boards are therefore not "after the chip". They are part of performance.

Priortech is exposed through Camtek and Access

Priortech is a holding company, so it should not be read like a regular operating equipment company. Its semiconductor exposure comes from two directions: Camtek, and Access through Amitec. Access is a China-based manufacturer of advanced substrates for the semiconductor industry, and it is what makes Priortech more than an indirect Camtek vehicle.

In its April 2026 presentation, Priortech presents 2025 as a record year for Access, with sales of about $291 million and improved profitability. Power Device reached 34% of Access sales, mainly because of components integrated into HPC computers used for data centers and AI applications. RF was still 44% of sales, but was affected by weakness in China and in mobile phones. Access backlog increased from about $70.0 million at the end of 2025 to about $83.6 million in mid-March 2026.

Priortech's value lies in the combination of assets, but so does the risk. This is a company with holding-company layers, parent-company debt, China exposure at Access, and a possible Shenzhen listing for Access that has not yet been completed. It belongs on the semiconductor map, but it needs to be read through holdings, corporate events and capital structure, not as a single-product operating company.

PCB Technologies expands in boards, substrates and microelectronics

PCB Technologies sits in boards, substrates and miniaturization. The company operates in three fields: printed circuit boards and substrates, electronics assembly, and electronic-system miniaturization in microelectronics. It defines substrates as multilayer materials that electrically connect semiconductors to printed circuit boards. The miniaturization activity includes development and production of substrates for the semiconductor industry and microelectronics assembly capabilities, including SiP.

In 2025, PCB Technologies reported consolidated revenue of $181.6 million, compared with $152.9 million in 2024. Gross profit was $38.0 million, operating profit was $13.7 million and annual profit was $13.1 million. Operating cash flow was $7.8 million. At the same time, the company ended the year with $3.8 million of cash and had used about $17.2 million out of $18.5 million of available credit facilities, meaning growth requires careful working-capital and financing management.

At the segment level, printed circuit boards and substrates generated $83.4 million of revenue in 2025, with backlog of about $56 million at year-end and about $60 million near publication. Electronic-system miniaturization grew to $8.6 million, compared with $4.5 million in 2024, with backlog of about $7 million at year-end and about $11 million near publication. This is still not the company's largest activity, but it is the clearest link to semiconductors and packaging.

PCB Technologies should be read in balance. There is a real link to the semiconductor value chain, mainly through substrates and miniaturization, but the company still relies on broader defense, medical and industrial activity. For semiconductors to become the central thesis, miniaturization and substrates need to keep growing and gain more weight inside the company.

Telsys Connects Investors To The Electronics Cycle

Telsys does not manufacture chips. It operates through two axes: electronic-component distribution, and Variscite, which develops and sells System on Modules. Its value-chain position is closer to modules, components and electronics customers than to chip production or semiconductor manufacturing equipment.

In 2025, Telsys reported revenue of $136.3 million, gross profit of $55.8 million and annual profit of $36.6 million. Variscite's SoM segment generated revenue of $80.3 million and segment profit of $39.3 million. Distribution generated revenue of $56.0 million and segment profit of $4.0 million. Variscite is therefore the main profit engine, while distribution links the company to the broader electronic-component cycle.

Customer concentration matters. Nvidia, through Mellanox and subcontractors, is an end customer responsible for more than 20% of consolidated revenue. Revenue from that customer reached $36.0 million in 2025, compared with $12.2 million in 2024. Distribution backlog that includes cancellation rights rose to $49.2 million at the end of 2025, but the company notes that customers can cancel or change part of the orders up to 45 days before delivery. Variscite's SoM backlog stood at $46.8 million for 2026 and $4.5 million for 2027-2028, compared with $29.0 million for 2025 and $3.1 million for 2026 at the end of 2024.

Telsys' chip link also runs through suppliers. Variscite relies on processors from leading manufacturers, mainly NXP, and since mid-2025 has faced a global memory-chip shortage influenced by AI infrastructure demand, which slightly affected gross margin. Telsys is not a semiconductor company, but it does feel the component cycle, memory shortages and the power of large end customers.

Arbe Faces The Test Of Commercial Production

Arbe is the most different company on the map. Technologically, it is one of the most chip-like names: a fabless developer of 4D radar chipsets for vehicles, using subcontracted manufacturing. But as a business, it is not comparable to Tower, Nova or Camtek. It is not a mature manufacturer and not an equipment supplier with a meaningful revenue base.

The latest financial information shows a company before material commercial revenue. In 2024, revenue was about $0.8 million and the annual loss was about $49.3 million. The company noted that it had not yet received significant production orders and had not transitioned into a production company, and that it planned to transition to production of its radar chip for customers during 2026, with no certainty that the timetable would be achieved.

The interesting side is the relationship with Tier 1 suppliers and automakers. The company notes that several Tier 1 suppliers are developing radar systems based on its chipsets, including Magna, Hirain, Weifu and Sensrad. There is also a preliminary Hirain order for 340,000 chipsets for the Chinese market, subject to becoming a binding order. If that happens, Arbe can move to a new stage. If not, it remains a development company that needs financing until commercialization.

Arbe therefore belongs on the semiconductor map, but not on the same shelf as profitable equipment companies. It is an option on technology adoption and automotive penetration. It should be read through binding orders, serial production, cash burn and access to capital, not through earnings multiples that are not yet relevant.

At Industrial Suppliers And Assembly Companies, Chips Are Not The Core Business

The final group matters because it prevents a common mistake: not every supplier to the semiconductor industry is a semiconductor company. Ashot, Imco and DSIT have real links to the chip world, but their center of gravity remains defense, aviation, assemblies or other systems.

At Ashot, semiconductor assemblies are absorbed inside defense activity

Ashot mainly operates in propulsion systems, gearboxes and complex assemblies for armored vehicles, aviation and industrial products. The semiconductor link comes through IDC, which manufactures complex assemblies for a leading international semiconductor customer, and through the PGV product line, a Personal Guided Vehicle used to carry wafers inside semiconductor fabs.

In its March 2026 presentation, the company showed 2025 operating profit of NIS 74.2 million, annual profit of NIS 56.8 million and EBITDA of NIS 92.7 million. It also presents backlog and framework agreements of about NIS 2.6 billion. But in the annual report, most aviation and complex-assembly backlog is not directly attributed to semiconductors, and the group is still mainly read through defense and aviation. Ashot is therefore an industrial supplier with a relevant semiconductor product, not a pure chip thesis.

At Imco, civilian electronics is small relative to defense

Imco supplies electrical, electromechanical, electronics and control systems. The semiconductor link is mainly in the civilian division, through PCBA and products for civilian customers, including customers in semiconductor manufacturing and testing equipment. But this is a small part of the company.

In 2025, Imco reported revenue of NIS 354.6 million, gross profit of NIS 86.3 million, operating profit of NIS 40.3 million and annual profit of NIS 21.2 million. The defense segment generated external revenue of NIS 304.6 million, compared with only NIS 50.0 million in the civilian segment. The Ministry of Defense represented 51.1% of group revenue and another government customer represented 21.2%.

In other words, even if Imco touches semiconductors through PCBA and equipment customers, it is mainly a defense and industrial-electronics company. To become a central semiconductor name, the civilian division would need to grow and become more material in consolidated results.

DSIT supplies assemblies for testing systems as a secondary activity

DSIT mainly operates in underwater defense systems and RT, or Real Time, assemblies. Within RT, it has activity in assemblies for testing systems used in chip-manufacturing processes. This is not completely immaterial: in 2025, revenue from assemblies for test systems in chip-manufacturing processes was $3.745 million, about 10% of company revenue, compared with $1.495 million and 7.1% in 2024.

The company as a whole reported 2025 revenue of $37.4 million, gross profit of $10.2 million and net profit attributable to shareholders of $5.6 million. The underwater defense segment represented $26.5 million of revenue, and the RT assemblies segment $9.9 million. The chip link exists and is measurable, but it is still part of a smaller segment within a company whose center of gravity is defense and underwater systems.

How To Read The Exposure Map Without Getting Misled

The important distinction is not company size, but business materiality. At Tower, Nova and Camtek, semiconductors are the business. At Qualitau and Priortech, semiconductors are a central growth engine, but with smaller scale or a holding-company layer. At PCB Technologies and CI Systems, the chip activity is measurable but not necessarily dominant. At Telsys, exposure runs through components, modules and electronics customers. At Arbe, commercialization still needs proof. At Ashot, Imco and DSIT, a real link exists, but it does not turn the whole company into a semiconductor company.

The Real Test For Local Semiconductor Companies In 2026

The TASE does not offer a semiconductor sector as broad as Nasdaq, but it does offer an interesting chain of exposures. At the direct end are Tower, Nova and Camtek. These are the companies where semiconductors are the core business, and industry-cycle changes can reach the financial statements relatively quickly. Behind them are Priortech and Qualitau, which provide more focused exposure but require careful reading of backlog, holdings, customers and corporate events.

The middle group includes PCB Technologies, Telsys and CI Systems. They have relevant activity, measurable numbers, and sometimes backlog or customers that connect them to the chip wave. But they should not be read the same way as Nova or Camtek. Here, the key is how much the activity weighs within the whole company, whether it is more profitable, and whether it is growing enough to change the company mix.

The industrial-supplier group includes Ashot, Imco and DSIT. They should not be removed from the map, but they also should not receive a semiconductor premium without evidence. A large order, a material customer or clear backlog from the chip industry could change their classification. Until then, they are mainly industrial and defense companies with real semiconductor points of contact.

The tracking question for 2026 is clear: whether the investment wave around AI, HBM, advanced packaging and data centers continues to turn into real orders at Camtek, Nova, Qualitau and Access. Whether Tower translates production volume into profitability. Whether PCB Technologies and CI Systems turn technology activity into more material growth. Whether Telsys maintains backlog and profitability despite component volatility. And whether Arbe moves from technology promise to binding production orders.

This is the working map. Not every company on it is a "semiconductor company". But each company on it helps explain a different part of the value chain, and that is what makes the local sector interesting: it is small, uneven, and deeper than it first appears.