TOKYO ELECTRON LIMITED

IR

Q1 FY2024 Earnings Release Conference Q&A

You said the total WFE*¹ market size is expected to be $200B for CY2024 and CY2025 combined, what proportion of this amount would each year have individually? Also, what will be the driving force for growth each year?

Due to the amount of time required to make adjustments in leading-edge logic/foundry inventories, customer investments are taking longer to recover than initially expected. Although there will definitely be a recovery over the years CY2024 and CY2025, we believe there will be more investment in CY2025. The greatest driving force will be servers. We expect server unit numbers to increase from CY2023 to CY2027 at a CAGR of around 8%. We also expect that PC and smartphone demand will recover due to an easing of inflation and increased IT investments by companies, as well as an increase in applications spurred on by the evolution of semiconductors. In addition, we believe that there will be steady investment in the "MAGIC*²" field, such as mature node semiconductors, etc. that are compatible with EV-related power devices and other various applications.

Regarding gross profit margin. Results for Q1 were 41.4%, and, after deductions, sales for Q2 would be 398.3B yen, resulting in a gross profit margin of 42.9%, an increase over Q1 according to the plan. What is driving this increase? Also, what factors will increase the gross profit margin for Q2?

The increase in gross profit margin for Q2 is primarily due to an increase in product profitability. We also expect that gross profit margin will improve further in Q2 due to increased sales.

You indicated that the sales composition ratio for China has increased, and that investment in mature nodes is particularly strong. Is this acceleration in investments by Chinese customers in Q1 due to the shift from leading-edge nodes to mature nodes? Furthermore, do you expect this trend to continue? In particular, has the growth of the China market already been factored into the strong investment recoveries for CY2024 and CY2025?

One reason for the acceleration in investments by Chinese customers is due to the demand for various semiconductors in a digital society. There is also less volatility in investments in mature nodes compared with investments in leading-edge nodes. In anticipation of export restrictions on some equipment by the governments of several countries, some Chinese customers are shifting their investments to mature nodes based on a stable growth strategy. In China, new customers are increasing, and investments in mature nodes are strong. We expect that this is not temporary, and that it will continue.

The stock market is demonstrating greater and greater expectations in response to the expansion of generative AI. What business opportunities are there for TEL, and which of TEL’s equipment do you expect will benefit most from this trend?

Focusing on AI-related servers, they represent approximately 8% of all servers as of CY2023, of which around 1% will be generative AI servers. Going forward, we expect that the AI server ratio will increase by around 2% points each year, and AI servers alone will contribute around $5B to the WFE market by CY2025. We expect that the spread of AI servers will result in increased business opportunities, such as in GPU and HBM*³, etc.
We benefit most in wafer bonders, and are already receiving orders. Front-end equipment will not see immediate demand for new equipment due to low utilization rates in customer fabs, but we expect benefits from next year and beyond.

How did you estimate that the WFE market size for AI servers would be around $5B in CY2025?

Our $5B estimate took into consideration factors such as number of server units, the proportion of servers for AI, and the ratio of generative AI among those servers.

Regarding the WFE market for AI servers in CY2025, what proportion of this market will be wafer bonders? What is your market share status, and what is valued about your wafer bonders?

We expect that sales will reach 2-digit billion yen in FY2024. We believe this market will grow rapidly in the future, and we expect to gain high market share. Leading customers in HBM*³ technology evaluate us highly in terms of wafer bonding precision and productivity. Given our broad product portfolio, we believe we can provide our customers with optimal solutions in both wafer bonders and surrounding processes.

As other etch equipment manufacturers do not have bonders, I think they collaborate with other bonder manufacturers. Will TEL be able to take advantage of the fact that you can provide both etchers and wafer bonders?

We believe that TEL does indeed have an advantage. We intend to demonstrate the value of TEL by proposing solutions that integrate front-end and back-end processes.

Cryogenic Etch: Will your competitors be able to double their etch rates by reducing wafer surface temperatures to -60℃ as opposed to -40℃, and changing the type of gas used as TEL has announced?

The hardware capable of cryogenic etch is significantly different from hardware for normal low-temperature etch, and is technically very difficult to achieve. This cannot be achieved by just changing the type of gas used.

Cryogenic Etch: It was my understanding that the maximum number of layers that can be etched at once in 3D NAND was 128 layers, but will cryogenic etch increase this layer count? Also, will we see 400-layer level devices in 2 to 3 years?

Cryogenic etch will enable etching more than 128 layers at once. As the number of layers increase, it is very advantageous for customers to be able to increase the number of layers they can etch at once, and thus reduce the number of etch steps. Channel hole etching requires an extremely large number of etch tools, and the market size for channel hole etching process will increase from roughly $0.5B in CY2023 to around $2B in CY2027. We believe this achievement will contribute to an increase in TEL's market share. Our expectation is that we will see 400-layer level devices adopted in 2 to 3 years.

I hear that the other SPE manufacturer is developing materials such as gases in addition to their equipment. Does TEL have any plans to develop materials in this way?

We at TEL engage in partnerships with gas manufacturers, selecting optimal gases during development. In our recent progress with cryogenic etch, we were able to take advantage of these partnerships to develop a new process using gases with extremely low carbon footprints that also enhance environmental performance.

WFE (Wafer Fab Equipment): The semiconductor production process is divided into front-end production, in which circuits are formed on wafers and inspected, and back-end production, in which wafers are cut into chips, assembled and inspected again. WFE refers to the production equipment used in front-end production and in wafer-level packaging production.

MAGIC (Metaverse, Autonomous mobility, Green energy, IoT & Information, Communications): A concept TEL has created to further expand business opportunities

HBM: High Bandwidth Memory

The above content is a summary of question and answers session.