Behind the shipment of High NA EUV lithography machine

Earlier, there was news that Intel will obtain the first High NA EUV lithography machine.

Meanwhile, industry body Trendforce reports that Intel will receive up to six of the 10 High NA ASML tools that may ship in 2024.

The article also quoted Samsung Vice Chairman Kyung Kye-hyun as saying that "Samsung has obtained priority for High NA device technology."

This seems to mean Intel gets the highest NA device, followed by Samsung due to its recent ASML announcement of a $755 million investment in South Korea.

This would put TSMC in an unusual third position, in stark contrast to its current dominance in EUV (70% of global EUV tools).

If we assume tool costs of $350 million to $400 million, and 2024 High NA sales for 10 tools will be in the range of $3.5 billion to $4 billion, even though not all revenue is likely to be recognized in the calendar year.

We believe the potential upside from ASML's High NA is not reflected in the share price, as ten tools sounds a bit exaggerated to most people, but if it does, it would mean strong Room for upside.

If we assume Intel does get the first 6 High NA tools as suggested, we can imagine Intel Portland will get at least the first two tools, with the others heading to Arizona and/or Ohio in 2024.

Of the other four tools, Samsung is likely to get at least two to three tools, while TSMC gets one or two tools.

We speculate that TSMC may push harder for multi-patterning of current EUV rather than jumping to High NA immediately.

Many in the industry believe that high numerical aperture EUV tools will be difficult to cost justify compared to existing multi-pattern EUV tools. This could be a smart move on TSMC's part, or it could be a mistake... time will tell. From Intel's perspective, they have no choice but to push hard because their past slowness on EUV is one of the reasons why TSMC has surpassed them in Moore's Law.

It was clear that Intel didn't want to repeat the mistakes of the original EUV tools, so it committed to ASML early on to get the first batch of High NA tools announced more than a year ago.

If High NA does succeed, this will be the leap Intel needs to catch up.

If our guess of 15 tools in 2025 is close, we'd expect the share to be more evenly split between Intel, Samsung, and TSMC, while Europe's IMEC, which works closely with ASML, will likely get one for R&D as well .

By 2026, there could be a 20-tool situation between the three major foundries/logic manufacturers, most likely adding a High NA tool for R&D to New York as recently announced.

We do not expect memory manufacturers to adopt high numerical aperture in the first few years as they are just now starting to adopt conventional EUV and are still 4 to 6 years behind foundry/logic demand for EUV lithography.

We wouldn't be surprised that after 3 years of High NA shipments, Intel owns most of the High NA tools, just like TSMC owns the lion's share of current EUV technology.

If Intel purchases 6 High NA tools in 2024, it means that High NA tool capital expenditures alone will reach $2.1 billion to $2.4 billion. It's a pretty big bet... but if it helps catch up with TSMC, it's money well spent. We have no choice because if we don't do this, Intel will fall behind TSMC and, if we're lucky, best case be on par with TSMC.

Although there are significant differences between EUV and high numerical aperture EUV tools, the basic concepts are the same. EUV is not a huge leap over ARF immersive technology.

Each device costs $350 million to $400 million, and chipmakers need to get these new tools up and running quickly.

They also have to focus (pardon the pun) on getting the financial return on single-pattern High NA versus multi-pattern standard EUV, which may not be a slam dunk at first, just like EUV will have a hard time proving multi-pattern ARF immersion (about Your cost advantage remains controversial).

We believe that if this news is true, or even close to true, it will be a major benefit for ASML and Intel. However, we don't see this as a negative for TSMC yet, as they are likely to maintain a strong position in the EUV space overall.

We believe ASML and its stock could experience a High NA EUV wave in 2024, assuming no production hiccups.

We further suggest that the recent changes to ASML's retiring CEO and CTO are unlikely to occur unless High NA EUV is safely on the road and any major technical challenges or other issues have been resolved.

While it may take some time for Intel to see the results of its High NA bet, we believe investors will be willing to overlook the negative impact of the spending required for High NA if the reward is regaining technology leadership.

*Disclaimer: This article is original by the author. The content of the article is the personal opinion of the author. The reprinting by Semiconductor Industry Watch is only to convey a different point of view. It does not mean that Semiconductor Industry Watch agrees or supports the view. If you have any objections, please contact Semiconductor Industry Watch.

Today is the 3625th issue of "Semiconductor Industry Observation" shared with you. Welcome to pay attention.