Good morning and thank you for attending SK hynix earnings release conference call. Today, we will begin with SK h ynix 's earnings presentation, followed by a Q&A session. Please press star one, that is star and one, if you have any questions. Please note that the presentation will be interpreted simultaneously, and Q&A session will be interpreted consecutively. With that, we are now ready to begin.
Good morning, afternoon, and evening. This is Park Seong-hwan, Head of IR at SK h ynix . Welcome to the SK h ynix 2026 third quarter earnings release conference call. Allow me to introduce the executives present here with me today. We are joined by President and CEO Kwak Noh-Jung, CFO Kim Woo-hyun, Head of DRAM Marketing Park Joondeok, Head of NAND Marketing Song Chang-seok, and Head of HBM Sales and Marketing Kim Kitae.
Let me issue a disclaimer that our third quarter results included in this conference call are consolidated figures and provisional in nature, as the external auditor's review has not yet been completed. Accordingly, they remain subject to change. In addition, forward-looking statements including market outlook and the company's plans may vary depending on changes in macroeconomic and market circumstances. With that, we will now begin SK h ynix earnings release conference call for the third quarter of 2026. Before we begin the earnings presentation, CEO Kwak Noh-Jung will first speak on the Solidigm IPO and shareholder returns. Following this, CFO Kim Woo-hyun will present third quarter earnings, the company's future plans, and market outlook.
Good morning, everyone. This is CEO Kwak Noh-Jung.
I would like to express my sincere gratitude to our shareholders for your unwavering interest and trust in SK h ynix , even amid the recent heightened market volatility. Today, I would like to speak directly to you about our decision regarding Solidigm's IPO, which has recently attracted considerable market attention, and a special dividend to share the returns on our strategic investment with our shareholders. First, Solidigm has been exploring financing options to support initiatives, such as expanding its manufacturing footprint to meet demand growth. Among the options considered, raising external capital through an IPO was expected to ease our investment burden and increase our capacity for shareholder returns.
However, after carefully considering recent financial market volatility and the correction in semiconductor stocks, Solidigm has concluded that an IPO would not fully deliver the anticipated benefits at this time and has therefore decided to discontinue its review of potential pre-IPO and IPO options. Solidigm will continue to strengthen its competitiveness and pursue growth opportunities in the AI data center storage market. Investments and financing arrangements to support these efforts will be pursued in a manner that enhances the corporate and shareholder value of both Solidigm and SK h ynix . We have upheld the principle of sharing the majority of the returns from our strategic investments with our shareholders. By using those returns to replenish our capacity for shareholder returns, which may temporarily decline as we invest for the future, we aim to protect our shareholders' interests while continuing to invest boldly in sustainable growth.
In line with this principle, we intend to return the gains realized from our investment in Kioxia this year to our shareholders through a special dividend. We will remain steadfast in this commitment, ensuring that our shareholders directly benefit from the success of our strategic investments. Our Chief Financial Officer will now present this quarter's results and outlook, along with the details of our shareholder return plan. Thank you.
Good morning, everyone. This is CFO Kim Woo-hyun. Allow me to first introduce the company's performance for the third quarter of 2026. In the third quarter, tight supply and demand conditions continued due to growing demand for AI memory, leading to sustained rise in memory prices. Both DRAM and NAND recorded high price increase rates for a second consecutive quarter, driven by continued price hikes centered on HBM, server DRAM, and enterprise SSD.
Consequently, third quarter revenue increased by 32% quarter-on-quarter and 328% year-on-year, reaching KRW 104.6 trillion. This marks the first time quarterly revenues surpassed KRW 100 trillion, setting a new record high. For DRAM, we expanded sales centered on AI servers, achieving sequential bit shipment increase of approximately 10% as planned. In particular, full-scale mass production and shipments of HBM4 commenced in the third quarter. Additionally, shipments of server DDR5 increased by over 40% compared to Q2, driving overall growth in DRAM bit shipments. With continued strength in conventional DRAM prices and an expanding sales mix of HBM4, which commands higher prices than the prior generation, DRAM ASP rose by high 20%, recording a high growth rate similar to the previous quarter.
For NAND, although bit shipments decreased slightly compared to Q2 as we adjusted sales to focus on high value-added products such as 321-layer TLC enterprise SSDs, ASP rose by mid-20% due to overall price strength and an expanded SSD mix. Despite unfavorable exchange rate effects due to the strong Korean won, product price increases and the expanded sales of high value-added products led to an operating profit of KRW 83.3 trillion for the third quarter. This represents a 38% increase quarter-on-quarter and a 632% increase year-on-year. Operating profit margin also improved by 3 percentage points quarter-on-quarter to reach 80%, setting new all-time quarterly records for both operating profit and operating profit margin. In the third quarter, depreciation and amortization expenses totaled KRW 4.3 trillion. EBITDA reached KRW 87.7 trillion with an EBITDA margin of 84%.
Regarding major non-operating items, foreign exchange related net loss of KRW 11.6 trillion due to exchange rates decline and valuation losses on investment assets of KRW 22.9 trillion were reflected, resulting in total net non-operating expenses of KRW 32.6 trillion. Consequently, income before tax was KRW 50.7 trillion, net income was KRW 39.8 trillion, and net profit margin was 38%. As of the end of the third quarter, cash and cash equivalents stood at KRW 114.5 trillion, an increase of KRW 26.5 trillion from the end of the previous quarter. Interest-bearing debt decreased by KRW 2.8 trillion -KRW 15.8 trillion. As a result, net cash expanded to KRW 98.7 trillion and debt to equity ratio improved by 2 percentage points compared to the end of the previous quarter to 5%. Next, I will discuss the market outlook.
The AI market is evolving endlessly as AI leads to productivity improvements and new revenue generation in various areas such as software development, customer response, and search. A virtuous cycle is forming where increased AI usage leads to further infrastructure investment. Such changes are evident in the scale of investment and the power capacity secured by major global cloud service providers. According to recent estimates from market research firms, global CSP CapEx is projected to grow by more than 60% in 2027, while secured power capacity is also expected to increase by over 100% year-over-year. More notably, this growth will not stop next year. The global AI data center infrastructure market is expected to grow at a high annual growth rate of 40% until 2030, driven by the commercialization of AI services and the expansion of application areas. This mid to long-term growth trend is expected to continue.
In particular, as AI plans complex tasks and carries them out over extended periods using multiple tools, AI compute requirements and the volume of data generated, transferred, and stored are rising sharply. Consequently, demand growth across the entire memory sector is significantly exceeding initial expectations. AI infrastructure investment is now moving beyond the initial phase, which relied on preemptive spending by a handful of players as participation broadens to include more enterprises and national governments. Accordingly, even allowing for some potential investment adjustments, we expect growth in AI infrastructure investment to remain robust. In the short term, enterprise and personal AI are expected to lead demand growth, while in the mid to long term, physical AI such as autonomous driving and robotics is expected to broaden the scope of demand.
While shipment volumes for PCs and mobile devices are expected to decrease due to cost burdens from rising memory prices and supply shortages, AI infrastructure demand growth is expected to more than offset this decline, sustaining growth in the overall memory market. On the other hand, memory supply cannot keep pace with this level of demand growth. Thus, we expect it to be difficult to fully meet customers' volume requirements in the near term. Given the complexities involved in implementing leading-edge processes and the lead times required to construct new production facilities, a situation where supply cannot keep up with demand is expected to persist for some time, a constraint projected to continue through 2030.
Despite the high level of demand, considering actual supply conditions, achievable demand growth for 2027 is expected to be at a similar level to this year, with growth rates of mid-20% for DRAM and high teen percent for NAND. We expect this favorable market environment to be a strong foundation supporting the company's earnings growth next year as well. Based on our expanded production capacity and competitiveness in high value-added products, we plan to achieve results at a significantly higher level, both in terms of absolute revenue and profit. Next, I will discuss the company sales plan for the fourth quarter and our future operational direction. In the fourth quarter, we plan to further accelerate supply expansion to respond more proactively to market demand. First, for DRAM, production increases due to the expansion of 1Z nm process and increased HBM4 production through M15X ramp-up will occur simultaneously.
Consequently, fourth quarter shipment volume is expected to grow meaningfully by mid 20% sequentially. For conventional DRAM, we are rapidly increasing the portion of 1Z nm products for server applications. In the fourth quarter, we plan to further strengthen our portfolio centered on high value-added server products by expanding sales of 1Z nm-based DDR5 and AI server LPDDR5 SOCAMM. Our shipments of 12-high HBM4 are proceeding smoothly according to the schedule agreed with our major customers. To actively respond to surging customer demand, we are accelerating M15X ramp and plan to reach full capacity by the end of the year. Accordingly, in the fourth quarter, HBM4 is expected to surpass HBM3E in both its share of total HBM shipment volume and revenue, establishing it as our flagship product. The company has finalized negotiations with major customers for next year, including HBM volume and prices.
These negotiations faithfully reflect the increased product value of HBM4. As a result, we expect the profitability gap between HBM and conventional DRAM to narrow meaningfully. For NAND, we plan to expand the portion of our flagship 321-layer process to over 50% of domestic capacity in the fourth quarter. Through this, we will expand the supply of 321-layer enterprise SSDs. Fourth quarter NAND bit shipments are expected to increase by mid-single percent sequentially. Building on the supply expansion, we plan to further broaden the role and business opportunities of NAND in the AI infrastructure market. We will expand the supply of high-performance, high-density enterprise SSDs suited for KV cache tiering. In the mid to long term, through HBF and system optimization technologies, we aim to evolve NAND into a core solution that enhances AI service throughput and cost efficiency.
To improve visibility into mid to long-term demand and secure a stable growth foundation, we are continuously expanding long-term agreements with customers. To date, we have signed LTAs with 14 key customers, including major global CSPs. While LTAs typically operate on a five-year basis, we are proactively discussing extension of existing contract periods with some customers. The fact that customers are requesting longer term collaboration based on our product roadmap and supply capabilities indicate that LTAs are evolving to strengthen customer relationships and business continuity. While pricing structures vary by customer and product, they are generally designed to reflect market conditions within a price range agreed upon with the customer. Furthermore, to ensure that customers' purchase plans lead to actual orders and shipments, we have included safeguards such as deposits.
This structure not only supports our stable profitability but also allows us to capture opportunities from market improvements by reflecting improvements in market prices and product value. The volume secured under LTAs currently account for more than 50% of our total capacity for 2027 and 2028, establishing a foundation for mid to long-term profit stability and demand visibility. Meanwhile, to respond to mid to long-term customer demand and growth opportunities, we are continuously pursuing investments to expand production capabilities. We expect Capital Expenditure in 2027 to increase significantly compared to 2026 as we invest in equipment to expand the share of leading-edge processes and ramp up our first fab in Yongin while domestic and overseas infrastructure investments accelerate. The scale and execution timing of investment will be determined by considering mid to long-term demand visibility secured through customer discussions, product competitiveness, profitability, and investment efficiency.
Finally, let me explain shareholder returns. Last August, we announced the shareholder return plan, including KRW 40 trillion of share acquisition and cancellation, and indicated that we would provide details on the size and method of additional shareholder return in the third quarter earnings call. Today, let me provide an update on the implementation of that plan and outline additional shareholder returns that reflect our improved business performance and investment results. The company has decided to pursue shareholder return of approximately KRW 100 trillion through the 2026 fiscal year. We plan to execute cash dividends of KRW 30 trillion or more, and share acquisition cancellations of KRW 60 trillion or more, including the KRW 40 trillion announced in August. Of this, the KRW 40 trillion share acquisition was completed last week and will be canceled shortly.
Through dividends, we will directly share the results of our earnings with shareholders and continuously acquire and cancel shares to ensure that the per-share value appropriately reflects our intrinsic value. Through this, we will fulfill our promise to return over 50% of free cash flow to shareholders. Today, the board of directors resolved on cash dividends of KRW 14 trillion and a share acquisition of KRW 20 trillion. Of this, treasury shares to be utilized for executive and employee compensation is expected to be approximately KRW 10 trillion. Excluding this, shareholder return for the third quarter is approximately KRW 24 trillion. I will now provide details on each. First, we are expanding cash dividends to reflect earnings improvement and investment results. We will determine dividend levels based on operating performance, and in cases where substantial investment gains occur, we will share those results additionally with shareholders through special dividends.
We will utilize the results not only from the growth of our core business, but also the realization of value from our invested assets as a source of shareholder returns. Accordingly, dividend for third quarter has been decided at KRW 20,000 per share, totaling KRW 14 trillion. This consists of a KRW 10,000 dividend reflecting earnings improvement and a KRW 10,000 special dividend to share the investment results from Kioxia. We plan to return most of the gains from our investment in Kioxia to shareholders through a special dividend. A portion will be paid as part of this quarter's dividend, with the remainder included in the year-end dividend once our financial results for FY 2026 have been finalized. Second, we will continue to acquire shares each quarter. The size of quarterly share acquisition will be determined through board resolutions that consider share price, market conditions, and cash flow.
This quarter, we plan to acquire a total of KRW 20 trillion of shares. Of this, KRW 10 trillion is acquired for shareholder return purposes will be canceled in full. The KRW 10 trillion acquired for executive and employee compensation purposes, as mentioned earlier, is not included in the 2026 fiscal year shareholder return amount. Moving forward, based on robust memory demand growth and the company's technological competitiveness, we will generate results in a stable manner and consistently share them with shareholders. Thank you. With that, we are now ready to take your questions.
Now Q&A session will begin. Please press star one, that is star and one, if you have any questions. Questions will be taken according to the order you have pressed the number star one. For cancellation, please press star two, that is star and two on your phone. The first question will be provided by Peter Lee from Citigroup. Please go ahead with your question.
[Non-English content]
Congratulations on the performance. My question is with regards to the AI investment. We see that the AI investment continues to grow, but then it seems as if the monetization following the AI demand or the AI infrastructure build-out is slow to be realized. As a result, there are some concerns in the market over the sustained AI demand. Against this backdrop, my question is, what is the company's basis for believing that the AI investment momentum will continue?
[ Non-English content]
Thank you for the question. We do understand that there is some discussion about pacing investment in AI, but we view this as part of the process where our customers are more closely examining the feasibility and profitability of individual projects and defining the timing and pace of investment. There are two main reasons behind our view that the AI investment momentum will continue.
[ Non-English content]
First, we are seeing customers' AI investment actually leading to greater adoption of AI services and producing business results.
[ Non-English content]
In their latest earnings release, major hyperscaler customers reported year-over-year growth of 30%-80% in their cloud revenue, a substantial portion of which we believe is coming from expansion of their AI business.
[ Non-English content]
We are seeing more long-term contracts for large-scale enterprise AI solutions as well, while cloud order backlogs have grown two-fold to five-fold YoY. Paying users are steadily increasing as well, confirming growth in major customers' AI-related sales and financial performance across multiple areas.
[ Non-English content]
Given this trend, major customers' AI sales and cash flow will only improve, which will also help ease the current market concerns on investment burden.
[ Non-English content]
At the same time, customers are working at various angles to ease current constraints, such as multi-site deployment, small or mid-sized data centers for which power is more readily available, and upgrading on-site power generation and cooling technology, which will help ease the investment burden.
[ Non-English content]
Second reason is that the investor base is becoming structurally more diverse.
[ Non-English content]
Early AI investment was concentrated among the big tech, so a single company adjusting its investment could directly result in overall demand volatility. But today, the investor base is rapidly broadening, also including non-tech companies investing in their own AI infrastructure, Sovereign AI, or on-prem demand. This marks a structural shift where investment adjustment by some would be offset by increased demand from others.
[ Non-English content]
For example, even if investment in some projects slows due to interest rate pressure, policy-backed Sovereign AI investment or a broader corporate AI adoption can still fill the gap. With a multilayered structure involving diverse participants with different motivations and resources, any single company's financial capacity or investment moderation affecting the overall investment trend is becoming limited.
[ Non-English content]
In sum, AI investment and demand intensity will remain sufficiently robust, supported by real business results and more diversified investment. Also, in the memory market, we expect the current structural supply-demand imbalance, where demand outpaces supply, to continue for the time being.
[ Non-English content]
[ Non-English content]
The following question will be presented by Sun-woo Kim from Meritz Securities. Please go ahead with your question.
[ Non-English content]
Thank you. I would also like to congratulate the company on the outstanding performance, and I would like to especially thank the CEO, Kwak No-Jung for the briefing on the company's update. My question is with regards to the shareholder return. What is the basis for the company to determine the proportion between the dividend payout and share acquisition out of the total shareholder return?
[ Non-English content]
Thank you very much for the question. For us, we sought to balance between cash dividend and share acquisition and cancellation, or meaning between the direct way of sharing our improved performance with shareholders and raising the value per share by reducing the number of outstanding shares.
[ Non-English content]
First, we decided on cash dividend at a level that will be sustainable. For this, we reflected our performance improvement as well as our historical dividend yield, global peers dividend yield, and appropriate level of cash holding.
[ Non-English content]
We also plan to share any meaningful return on our investment when it arises, mainly the return on our investment in Kioxia. Most of it will be paid out in special dividend so that it can be actually felt by our shareholders. Although we have yet to close the books for this year, we decided that it would be meaningful to share at least part of the expected returns with our shareholders first. So part of the special dividend will be paid out early in the third quarter, and the rest will be included in the year-end dividend after finalizing the return on investment.
[ Non-English content]
After the cash dividend, the rest of the total shareholder return is allocated to share acquisition. We believe our share price today is undervalued, not adequately reflecting our earnings and cash flow. That is why we decided that share acquisition and cancellation would be an effective way of raising the per share value. The acquisition will continue every quarter within the annual size, and the size of the quarterly acquisition will be determined by the board of directors considering such factors as the share price, valuation, cash flow, and so forth.
[ Non-English content]
Going forward, we will faithfully implement our shareholder return policy based on our robust cash generation, and we remain committed to continue to enhance our shareholder value.
[ Non-English content]
Thank you very much. We will take the next question.
[ Non-English content]
The following question will be presented by Giuni Lee from Goldman Sachs. Please go ahead with your question.
[ Non-English content]
Congratulations on the outstanding performance. My question is on the LTA. It was explained earlier that currently the volume secured through the LTAs is about 50% of the volume of the total capacity for 2027 and 2028. What would it mean for the company's performance and potential for increment growth in the longer term?
[ Non-English content]
Thank you for the question. By increasing LTA coverage, our goal is to improve longer-term demand visibility, which can lead to stable revenue and earnings growth.
[ Non-English content]
The volume already secured under LTAs and the demand currently being discussed with customers taken together means that our DRAM and HBM volume for the next three years through 2029 are effectively sold out.
[ Non-English content]
The volumes under discussion are based on customers' concrete purchasing plan and are highly likely to come through. Overall, we believe our medium to long-term sales visibility has significantly improved.
[ Non-English content]
As mentioned earlier, the volumes committed under LTAs account for more than 50% of total Capacity in 2027 and 2028, and we plan to keep it similar over the medium to longer term.
[ Non-English content]
This arrangement will help us secure a reliable earning space. At the same time, we can utilize the non-LTA Capacity to flexibly respond to market conditions and demand for next generation products to keep strengthening our future competitiveness.
[ Non-English content]
Thank you. Next question please.
[ Non-English content]
The following question will be presented by Simon Woo from Bank of America. Please go ahead with your question.
[ Non-English content]
My question is with regards to possible overproduction, and that is because there have been some challenging cycles in the past. The question or the concerns over possible overproduction come from the fact that because of the very strong market demand that is continuing, h ynix continues to build either new fabs or build out or upgrade existing fabs, including the ones in Yongin. What if demand were to fall short of expectations, then what would be the company's strategy or the management's plan against it? I would also like to congratulate the company on the excellent performance as well as the very detailed shareholder return plan.
[ Non-English content]
Thank you for the question. It is true that the company is currently pursuing medium to long term investment to accommodate the strong demand and secure enough space for the equipment and production needed for tech migration. But more investment or more fabs does not necessarily mean more supply.
[ Non-English content]
With the rising share of high-performance products like HBM and increasingly complex processes, more resources are needed for production. But increasing bit output with finer technology is becoming more and more limited. Our investment today is not only to expand production capacity, it is also intended to build the production base that is needed for a tech migration and expansion of high value-add product lineup.
[ Non-English content]
Our investment will be executed in stages. The foundational infrastructure will be readied in advance for medium to long term demand, but the key resources needed to increase supply, such as equipment moving and mass production ramp-up, will be adjusted in line with demand. For example, for new fabs, including the Yongin cluster and M17, we plan to accelerate construction of the shell structure, but clean rooms will be opened in phases. This approach is specifically intended to allow flexible operation in the future.
[ Non-English content]
When assessing demand, we don't just look at the customer's CapEx or demand forecast. We also review their financial health and funding capability, as well as the readiness of key supply chains before reflecting this in our investment plan.
[ Non-English content]
Our own financial soundness is also an important parameter. While maintaining sufficient financial room in light of our medium to long term net cash target and appropriate cash holding, we will manage the scale and pace of investment based on demand visibility and investment returns.
[ Non-English content]
To sum up, we will prepare our medium to long term supply infrastructure as planned, based on improved demand visibility. Actual production capacity will be ramped up only in line with demand and profitability. If conditions do differ from our expectations, we will make sure to respond to any risks by flexibly adjusting our investment disbursement, production plan, and resource allocation.
[ Non-English content]
Thank you. We will take the next question.
[ Non-English content]
The following question will be presented by Jay Kwon from JPMorgan. Please go ahead with your question.
[ Non-English content]
Congratulations on the good performance and thank you very much for taking my question. The HBM4 supply was mentioned earlier, but I would like to get an update from the company on the specifics of the HBM4 supply. Can you also share your plan for the follow-up, the HBM4E and custom HBM?
[ Non-English content]
Thank you very much for the question. As mentioned earlier, HBM4 is expected to crossover HBM3E in the fourth quarter in both bit shipment and revenue share, thanks to mass production proceeding without disruption.
[ Non-English content]
The yield is also quite stable. The company is maintaining industry best level yield at both our fabs and back-end processes. In particular, our back-end yield refers to the entire process, not just individual steps like bump, stack, or KGSD test. On this basis, HBM4 already exceeded its planned yield, so the target was revised up. HBM3E has also surpassed the planned mature yield.
[ Non-English content]
For now, we are focused on maximizing productivity to meet strong customer demand. Our performance in the HBM business will once again demonstrate our unparalleled competitiveness.
[ Non-English content]
Of course, we are also steadily preparing for next generation products. For example, for the HBM4E, we already completed sample shipment in the second quarter. Development and qualification are moving ahead in line with customers' timeline.
[ Non-English content]
And looking further ahead, in addition to stable supply of standard HBM, we are seeing growing importance of custom HBM, which is optimized for each customer's system and workloads. So we are working closely with major customers on related projects.
[ Non-English content]
As the first supplier to our key customers, we will keep growing our HBM business, drawing on proven product competitiveness and reliable mass production capability.
[ Non-English content]
Thank you. Next question.
[ Non-English content]
The following question will be presented by Dong-h ee Han from SK Securities. Please go ahead with your question.
[ Non-English content]
Thank you for taking my question, which is also on the HBM, particularly on the possibility of lowering the HBM stacks. What is the company's view on this possibility of having shorter HBM stack? What would be the potential impact on the HBM business, including HBM shipment and profitability?
[ Non-English content]
Thank you for the question. The recent moves by some customers to adjust HBM specifications illustrate that demand is indeed overwhelming supply. With HBM supply remaining constrained, customers need to build more AI accelerators and systems using limited quantity. We see the stack lowering as a strategy to increase the number of cubes that can be obtained for the same capacity.
[Non-English content]
Having said that, lower stack does not necessarily mean a proportionate decrease in HBM capacity. Even if it goes from 12-stack HBM4 to eight-stack HBM4E, the increase in capacity per die can offset a substantial part of the impact.
[ Non-English content]
Even when considering such adjustment, our medium to long-term HBM demand outlook remains unchanged. Demand remains so strong that customers' volume requirement is more than double that of what we can supply today, which is also indicative of customers' continued strong preference for our products.
[ Non-English content]
Neither the supply side will be affected. Supplying the same bit volume with lower stacks requires more cubes, which in turn increases the need for base dies and back-end processing. We already incorporated this into our production and supply plan, and we will implement our HBM shipment and supply as planned.
[ Non-English content]
As for profitability, it is not determined solely by the stack count. There are a combination of factors like terms of trade for each product mix, yield, back-end productivity, and cost. Next year, share of HBM4 and its price will increase while mass production yield and cost competitiveness will further improve, making it likely that HBM revenue and profit grow significantly year-over-year, with meaningful improvement in profitability as well.
[ Non-English content]
Looking ahead, profitability is expected to remain high, far above the peak in past cycles.
[ Non-English content]
Thank you. Next question.
[ Non-English content]
The following question will be presented by Young-gun Kim from Mirae Asset Securities. Please go ahead with your question.
[ Non-English content]
Congratulations on the record performance. My question is on the 4F². We see that other companies are also working on new cell structures and next-generation DRAM technology, which is heightening interest in long-term technology competition. Is there an update on the preparation for the 4F², and what is the company's areas of development focus to strengthen our technological competitiveness?
[ Non-English content]
Thank you for the question. The 4F² is a next-generation DRAM technology that increases density with a smaller cell area than 6F², and the applicability and strategy would depend on each company's scaling capability, process environment, and cost structure.
[ Non-English content]
For the company, our development plan is to utilize the 6F² structure to the extent possible through process maturity and stable tech platform, then transition to 4F² or 3D DRAM platform.
[ Non-English content]
Through the EUV-based scaling and process innovation, we have secured the technological ability to further extend our current 6F² platform. At this point, we believe that the most efficient strategy for the company is to strengthen our product competitiveness by making the most of a platform whose mass production reliability and cost competitiveness have already been proven.
[ Non-English content]
To apply a new cell structure into the products, we must secure not only performance, but also yield, reliability, and cost competitiveness. Not only that, the product design and customer qualification must also be completed. This is why the timing of technological development and the timing of commercialization are distinct from each other.
[ Non-English content]
We are preparing a step-by-step roadmap covering from the extreme scaling of the existing DRAM to 4F² and 3D DRAM. Each technology will be applied to the right product at the right time. The right time, meaning when it is ready technologically and economically. We will keep drawing on our process and design capability, production experience, and customer cooperation, and sustain our technological competitiveness in next generation DRAM by combining performance, cost efficiency, and producibility.
[ Non-English content]
Thank you. We will take the last question.
[ Non-English content]
The last question will be presented by Young-ho Ryu from NH Investment & Securities. Please go ahead with your question.
[ Non-English content]
Thank you for taking my question, which is on NAND. Now, we see these days that the size of KV cache is growing rapidly, and now there is also work underway to store and reuse it across multiple memory tiers. How does the company expect this to affect future NAND demand, and what type of preparation is being made?
[ Non-English content]
Thank you for the question on NAND. As agentic AI and long context use cases spread, the size of KV cache is growing rapidly, making it difficult to store it all in HBM or DRAM due to capacity and cost constraints. As a result, a tiered memory architecture is becoming increasingly important, which distributes data across HBM, DRAM and SSD according to access frequency.
[ Non-English content]
This is expected to create new demand for NAND to store and reuse the context needed for AI inference. SSD is already being used in some AI systems to offload KV cache, and we believe its role will expand into the context memory tier that allows multiple servers and GPUs to share large scale KV cache. In response to this opportunity, the company is developing memory tiering technologies, including the semantic aware lifecycle tiering for KV and CMM-Hybrid. The SALT-KV is a software technology that optimizes data allocation across HBM, DRAM, and SSD by assessing the reuse value and storage cost of each KV cache. CMM-Hybrid is a CXL based hybrid memory technology that stores the entire KV cache on SSD. It then predicts which data is likely to be needed and preloads it into DRAM.
[Non-English content]
And more recently, through system demonstrations based on actual AI workloads, we are defining the applicability and performance requirements for eSSDs in real inference environment to make it ready for adoption in our customers and platform partners architecture. And for the medium to longer term, we plan to address the new memory tier between HBM and SSD using HBF, which combines high bandwidth with NAND's high capacity and cost efficiency. By preempting the new NAND demand opportunities arising from changes in AI system architecture, we will sustain growth in our NAND business.
[ Non-English content]
Thank you for the answers. And due to the time constraint, this brings us to the end of the Q&A session. We will conclude the SK h ynix third quarter 2026 earnings release conference call with the CEO's closing comments.
[Non-English content]
Now, we have delivered record results for six consecutive quarters.
But what we find more significant is that the underlying business environment is changing. The customer trust that we have built through our technological leadership is translating into long-term agreements, giving us greater medium to long-term demand visibility and a stronger foundation for sustained profitability. This means that the potential for a new phase of growth that is distinct from the past is manifesting through concrete changes in our business structure. Our task now is to turn these changes into lasting competitive advantage. We will extend our current leadership in memory, including in HBM, to next generation products and become the first partner that customers turn to as they prepare for the future. We will translate market opportunities into results and further strengthen our ability to sustain that performance. And as our business grows and our investments deliver returns, the rewards we share with our shareholders should also increase.
The shareholder return plan announced today reflects this principle. We will continue to build confidence in our future growth through sustained performance and by sharing the rewards with our shareholders. Thank you very much for your continued interest and support. That concludes the closing remarks and the earnings release conference call. Thank you very much.