Q2 FY2024 Earnings Release Conference Q&A

Although we are still at the stage of consuming inventory for DRAM and logic/foundry, we expect that there will be a significant recovery from the middle of CY2024 onward, and that there will be further growth in CY2025 driven by AI server demand. On the other hand, we expect that adjustments for NAND will continue until the end of CY2024.

We made an upward adjustment of $15B for WFE in CY2023 compared to the previous market outlook and one factor for this increase was due to a delay in the delivery of lithography equipment and other equipment causing a delay in sales recording. We expect that our SAM will expand through an intrinsic recovery in the market in CY2024. In addition, the reason we were able to end up with better results than our initial estimate for the first half of FY2024 was due to some results from the second half being recorded ahead of schedule. The market is trending toward recovery overall, and the portion that was brought forward is being covered by market recovery and investment in China.

This upward revision of $15B in CY2023 WFE was due largely to portions that did not correspond to our SAM. It is not a positive situation for our market share at present, due in part to the depreciation of the yen. On the other hand, we are steadily increasing the number of PORs*³, and we believe that we will increase our share in CY2024 and CY2025 as demand recovers.

We were able to achieve a higher margin in the first half than we expected, by selling high value-added products at the appropriate price. We expect gross profit margin to improve in the second half as sales will be higher. We have seen no cases of sharp increases in purchase prices from suppliers this fiscal year. As technological innovation is extremely fast in this industry, customers request for new added value every 1.5 to 2 years. By listening to the voices of suppliers and selling products with new added value at appropriate prices, we believe we can continue to maintain sound business.

Compared to conventional technology, cryogenic etching has superior productivity and environmental performance; TEL has already acquired development PORs from some customers, and evaluations are progressing steadily. As the 400-layer level is our main target, we expect the prime timing for investment will arrive in CY2025. This technology may be used ahead of schedule and applied to previous nodes, and we are working with customers to evaluate it further. Although we do face competition, our differentiated technology has been recognized by customers, and our analysis suggests that we maintain the advantages we have previously announced.
Regarding cleaning equipment, while other competitors also have good equipment, we are expanding our market share by achieving greater technological differentiation. We are very hopeful for the future.

Though this fiscal year has been an adjustment period, we have managed to achieve our initial plans with comparatively strong accuracy. The performance of coater/developer was particularly strong compared to our initial expectations. There are no particular risks we need to be mindful of for FY2025. It has also been notable that orders for HBM*⁴ bonders for mass production use have finally begun to accelerate abruptly. Additionally, we were able to acquire PORs in conductor etching for logic/foundry patterning.

In China, the segment we call "MAGIC*⁵" is the primary market. We have seen around 20 to 30 new customers, and going forward we expect to see the Chinese market grow even further. When it comes to the global WFE market for MAGIC applications, we expect it will be around $30B for CY2023, moving forward to CY2030 we expect this to reach around $50B or even greater.

We have already received inquiries from China for CY2024, so we can expect some visibility. Our forecast for the first half of CY2024 in particular shows that China will continue to represent around 40% of sales by region. When we say that CY2025 is going to be a big market, this isn't just because of the Chinese market: we expect that AI servers will be a driving force with a CAGR of 31% from CY2023 to CY2027. Inventory adjustment will be completed by then, and we also expect to see further growth from NAND in addition to the DRAM as well as logic/foundry. Also, in addition to demand for PC and smartphone replacements, we have strong expectations for the addition of new on-device AI features on those devices.

We are not in a position to comment on regulations. At the same time, we expect that demand will continue to expand due to the realization both digitalization and decarbonization, the expansion of the semiconductor device market, and an increase in new applications. We believe these demands will be covered somewhere regardless of the manufacturing location. We also believe that WFE overall will grow sustainably from a mid- to long-term perspective.

As inquiries have increased rapidly from South Korean customers in particular, things were quite tight for us at the beginning of the term. Since then, however, we have been able to expand our production capacity by several degrees thanks to the cooperation of our partner companies. Our inability to meet HBM demand due to a lack of parts and components has already been resolved.

Besides bonders, we are seeing various business opportunities in generative AI and are preparing new products as well. We are currently devising strategies to maximize business opportunities for TEL, and will announce them when the time comes.