ASML Holding (ASML 2.04%) is often considered a linchpin of the semiconductor market for two reasons. First, the Dutch company is the world's largest manufacturer of lithography systems, which are used to optically etch circuit patterns onto silicon wafers. Second, it's the only producer of extreme ultraviolet (EUV) lithography systems, which are required to manufacture the world's smallest, densest, and most power-efficient chips.
That's why ASML controls more than 90% of the global lithography market. It also took ASML more than three decades to perfect its EUV technology. To many investors, ASML's early mover's advantage and market dominance suggest it only needs to worry about macro and regulatory headwinds instead of competitive challenges.
However, Canon (CAJ 3.72%) -- the Japanese company best known for its cameras, printers, optical products, and medical equipment -- recently challenged that notion with a new "nanoimprint" semiconductor manufacturing system called the FPA-1200NZ2C. It claims this new system can produce the equivalent of 5nm to 2nm chips by imprinting circuit patterns like a stamp instead of etching them with lasers. Could this new chipmaking system turn Canon into the next ASML?
Image source: Getty Images.
What drove Canon to challenge ASML?
Canon already produces lower-end deep ultraviolet (DUV) lithography systems for making older and less complex chips, but those systems only accounted for a single-digit share of the global lithography market. It sells those systems through its industrial division, which only accounted for 7% of its sales in the first half of 2023.
Canon's industrial division sells lithography systems for manufacturing semiconductors and flat panel displays. In the first half of the year, the segment's sales declined 7% year over year -- making it Canon's worst-performing business unit -- as its soft sales of flat panel display systems offset its growth in semiconductor systems.
On the bright side, Canon expects its semiconductor lithography systems business to keep growing as the expansion of the generative artificial intelligence (AI) market drives data centers to buy more chips. It also expects its total semiconductor lithography system shipments to rise 11% to 195 units this year. Therefore, it seems like an ideal time for Canon to expand the semiconductor lithography segment with new products -- but it can't possibly catch up to ASML in the EUV race.
To address that issue, Canon has been developing a nanoimprint lithography (NIL) technology as an alternative to EUV since 2004. However, the adoption rate for NIL systems has been sluggish due to ongoing concerns about defects during the manufacturing process. Unlike EUV systems, which optically etch circuit patterns without touching the wafer, NIL systems need the template to touch the wafer to imprint their designs. The tiny defects which could emerge from that physical contact don't matter too much for larger chips, but they could adversely affect smaller chips.
Proponents of NIL systems believe its defects will be resolved as the technology improves. Canon's commercial launch of its first NIL system suggests that moment has finally arrived.
A reluctancy to use those systems
Canon's launch of its NIL system caused ASML's stock to swoon briefly, but I don't think it represents a near-term threat to ASML. ASML's top customers -- Taiwan Semiconductor Manufacturing, Samsung, and Intel -- won't abruptly switch from EUV to NIL systems. Those leading foundries are set up to run EUV systems, and their fabless customers expect their chips to be manufactured by ASML's industry standard systems instead of untested NIL systems. Any adoption of NIL systems will likely be done in very small doses to prove the technology is viable and won't cause defects during the manufacturing process.
Canon's NIL system might represent a longer-term threat to ASML if more foundries take the technology seriously. It could also gain ground if its system costs a lot less than ASML's EUV systems -- which cost about $200 million apiece -- but Canon hasn't revealed its final price yet.
I don't think ASML will simply sit still and wait for that to happen. Instead, it will likely focus on rolling out its newest high-NA EUV systems to produce sub-2nm chips. Those systems should enable it to maintain its dominance of the lithography market and discourage its top customers from dabbling with Canon's NIL systems.
Canon won't become the next ASML
Canon might be an interesting play for investors who want some diversified exposure to the printing, camera, medical device, and lithography sectors. From 2022 to 2025, analysts expect its revenue to grow at a compound annual growth rate (CAGR) of 3%. However, ASML expects its revenue to grow at a CAGR of 10% to 14% from 2022 to 2030. That confident forecast implies it will continue to dominate the lithography market as it continues to lock in the world's top foundries.
Canon's announcement is a reminder that ASML isn't alone in the lithography market, but it's not a meaningful threat to its long-term growth. Canon also certainly won't become the next ASML, but that's really an apples-to-oranges comparison. The former is still mainly a printer and camera maker, while the latter is a pure play lithography leader.
