Good morning. Allow me to first report to you the company's performance in the second quarter of 2022. The second quarter saw rising global inflation and the persisting Russia-Ukraine war. Shutdowns in some regions of China and continuing supply chain issues prolonged the difficult business environment. The rising inflationary pressures and deepening recession concerns towards the end of the quarter affected not only consumer confidence, but investment sentiment as well, further dampening the IT demand. All of this led to DRAM bit shipment growth of around 10% in the second quarter compared to the previous guidance of mid-teens %. Demand for PC and mobile softened significantly, but sales grew for a computing product for which demand was relatively solid. ASP fell low single-digit % from the previous quarter. NAND bit shipment, including Solidigm, grew by high single-digit % compared to the guidance of over 20%.
The company tried to make up for the soft demand in consumer products with increased sales of SSD products. Excluding Solidigm, NAND bit shipment growth was low teens %. ASP rose by low single-digit %, both including and excluding Solidigm. Despite shipment growth falling somewhat lower than the guidance, the company's revenue in the second quarter was KRW 13.81 trillion, up 14% from the previous quarter and 34% year-on-year, the highest ever in a quarter. This is thanks to positive shipment growth and the strong dollar offsetting the decline in DRAM ASP. The recently acquired Solidigm made a contribution to achieving record high revenue, but even excluding Solidigm, the second quarter's revenue in Korean won terms is the highest ever. This was made possible by the unsparing hard work by all SK hynix members, despite the mounting adverse conditions like lockdowns in China.
Operating profit in the second quarter was KRW 4.19 trillion, up by 47% from the previous quarter and 56% year-on-year. Operating margin was 30%, improving from 24% in the previous quarter. Despite the rise in raw material prices, unit cost declined for both DRAM and NAND as a result of yield improvement, as well as the increase of our 1α nm DRAM and 176-layer NAND in the mix. Therefore, profitability of both DRAM and NAND improved sequentially. Depreciation and amortization in the second quarter was KRW 3.48 trillion, increasing by approximately KRW 0.83 trillion quarter-on-quarter. EBITDA in the second quarter was KRW 7.68 trillion, and EBITDA margin 56%.
There was net interest expense of KRW 89.8 billion, as well as foreign currency translation loss of KRW 311.8 billion from foreign currency denominated debts due to the strong US dollar and on Kioxia investment due to the weak Japanese yen. There was also a gain of KRW 124.1 billion from the sale of the company's Bundang office building. Including all these non-operating gains and losses, second quarter's net non-operating loss was KRW 225.9 million. Income before tax was KRW 3.97 trillion, with corporate tax expense of KRW 1.09 trillion. Net profit was KRW 2.88 trillion, with net profit margin of 21%.
Consolidated cash balance at the end of Q2 was KRW 7.5 trillion, down by KRW 0.5 trillion from the previous quarter. Interest-bearing debt was KRW 19.38 trillion, up KRW 1.27 trillion quarter-on-quarter. Debt to equity ratio and net debt to equity ratio at the end of Q2 was 29% and 18% each, flat or up slightly from the previous quarter, respectively. Next is the company's market outlook and plans. In the past two years, IT demand growth exceeded expectations due to the increase in online activities in the pandemic, followed by pent-up demand. Global supply chain disruptions held back product builds. Supply not being able to meet the end consumption demand, which also negatively affected demand for memory products to some extent.
The recent inflationary pressures and concerns of recession are quickly dampening consumer confidence, with businesses also noticeably starting to tighten costs. Although supply chain issues lasting into the first half have begun to dissipate, we are now facing a slowdown in demand, compelling us to considerably adjust in memory demand outlook for the second half. Demand for consumer products like PC and smartphones, which remained soft in the first half, will be adjusted significantly for the year as they are directly affected by the weakening purchasing power due to inflation. Demand for enterprise PCs, initially expected to be strong in the second half, is falling short of expectations, with annual PC shipment now likely to fall year-on-year. On the back of mixed increase towards high density PCs such as gaming PC, memory contents are expected to grow by more than double digits, offsetting the PC shipment decline.
Smartphone shipment is also projected to fall year-on-year, with a sharp slowdown in demand for low to mid-end products, which is a segment highly sensitive to economic conditions. However, in the second half, solid demand for flagship models is expected with new product launches which can improve mobile demand to a certain degree. Although the set build was impacted by the shortage of some parts, server demand remained solid in the first half. In the second half, it is expected that customers may adjust inventory on hand first as businesses begin to cut costs and reduce investments out of recession concerns. However, as individual enterprises reduce on-premise investments, it is highly likely for them to increasingly turn to using cloud services.
In other words, they are expected to increase OpEx instead of CapEx in times of high uncertainties, and thus will sustain the cloud industry's growth over the longer term, as well as memory demand for data centers. Given the market uncertainty, market growth for the year is expected to be a lot lower than our expectations earlier in the year. Thus, the company's plan for third quarter bit shipment will also be revised down compared to the previous plan. Meanwhile, the company's cumulative investment in the first half in terms of equipment installation base was KRW 8.8 trillion. Per the guidance last quarter, CapEx for the year will increase year-on-year.
At the same time, the company is carefully reviewing various scenarios of the production volume and the necessary CapEx to meet next year's demand, considering the level of inventory that we may end up having at the end of this year. We are also having close discussions with our major customers to better estimate next year's demand situation, and will reflect the outcome of our discussion in setting next year's CapEx plan. We are in a period where it is difficult to achieve supply flexibility due to lingering supply chain issues, while macroeconomic and market demand uncertainties are ever increasing. The company believes that at a time like this, we need to focus on our core business competitiveness to prepare for the time of demand recovery and lead growth in the memory industry over the long term. Next, let me report to you the company's technology and product roadmap.
Our 1α nm DRAM and 176-layer NAND are showing smooth yield improvement and are increasing in the production mix. The company will minimize the impact of equipment delivery delays on the tech migration timeline to meet the targets for leading technology process mix by year-end and continue to fulfill customer demand. The company's industry leading 1α nm process will be expanded to all applications in the second half. We will also focus on increasing density, such as 12-gigabit and 16-gigabit LPDDR5 and 16-gigabit and 24-gigabit DDR5 products. DDR5 products are provided mostly for PC customers this year, while for servers it will be more from the next year due to the delayed launch of supporting CPUs.
However, the portion of mobile LPDDR5 will increase sharply in the second half as it is adopted in new flagship models and will take up over half of all mobile solution products, excluding LTE. In particular, high density mobile solution based on LPDDR5X to be adopted in the second half is a high bandwidth product that provides higher performance than the previous tech generation, offering 33% higher speed and 25% lower power consumption. I would like to highlight that this solution can enable our mobile customers to differentiate in hardware specifications. Furthermore, the company's world's first HBM3 went into mass production in June and began to be supplied to Nvidia, our major graphic memory customer. The high bandwidth product will be combined with Nvidia's H100 to be used in the latest AI-based technology like accelerated computing.
The company will further boost our industry-leading performance and quality that we have been maintaining since HBM2E to firmly establish our position as the number one provider in the fast-growing HBM market and lead the high-performance premium memory market. Despite increasing macroeconomic uncertainties and rapid changes in demand, the company remains convinced of the long-term growth potential of the industry. Although it is hard to predict and prepare for the many variables that affect demand, the company will strive to sustain growth based on our business management capabilities that have grown stronger over the years. Last, let me brief you on the company's ESG activities and performance in the last quarter.
As a member of the Responsible Minerals Initiative, or the RMI, the company is trying to end human rights violations and environmental degradation that occur in the process of mining minerals in conflict-affected or high-risk areas. To ensure responsible sourcing of minerals, the company has completed due diligence on suppliers of raw materials that contain four conflict minerals called 3TG, tantalum, tin, tungsten, and gold according to the OECD guidance. We have undertaken improvement actions wherever necessary and achieved 100% responsible minerals assurance process certification in all in our supply chain. Also, the list of 227 smelters has been posted on the company's website. The company will go beyond conflict minerals to ensure responsible mineral sourcing and will transparently disclose related information.
The company recognizes that risks stemming from the supply chain can seriously affect not only the company, but also the industry as a whole, and intends to thoroughly address them. We will keep communicating and sharing our ESG values with our suppliers in order to build together with our partners a sustainable semiconductor ecosystem for shared growth. Thank you. With that, we're 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. In order to allow as many Q&A chances as possible within the restricted time, we would appreciate only two questions per each participant. The first question will be provided by J.J. Park from JPMorgan. Please go ahead with your question.
Yeah.
Now my first question is regarding bit growth. You'd mentioned that the company has mentioned that in the third quarter, the company will revise down the bit growth guidance. The question is then what is going to be the bit growth guidance for the third quarter? Also, for the year, the industry's bit growth is going to be around low teens%, as has been forecast by the company. Does the company also believe that the company's bit growth for the year will be roughly similar to the industry average? Actually, the second question is about the inventory policy. In the second quarter, the bit shipment has fallen under the guidance for both DRAM and NAND.
As a result, it is expected that the inventory level would be higher than it used to be. What is the company's inventory management policy into the second half of the year? Does the company foresee the inventory being carried into next year? Or is there some kind of threshold that the company has, whereupon if it is reached, then the company plans to sell down the inventory?
Thank you for the questions. Now, as everyone knows, when we look at the macroeconomic conditions, then they still remain quite murky into the second half of the year. Likewise for the memory economy that is highly correlated with the macroeconomy, so we all know that the situation does not bode well for the memory industry at this time.
Because of this, compared to the second quarter, in the third quarter, we expect the both for DRAM and NAND to be a flat QOQ. Please understand that it is very difficult for us to specify any numbers at this time because of the very high uncertainties. Because of that uncertainties, we are seeing numbers change almost on a day-to-day basis. Internally, we are planning to be at least flat QOQ in the third quarter. For the third quarter, in terms of the bit shipment as well, the target has been also adjusted downward as well. It is also planned to be flat.
Of course, despite the uncertainties for the market on the whole, we believe that the market will grow by low teens% for DRAM, and the company's target is also aligned with the market growth. For NAND, the NAND numbers will also be much lower than initially expected. Now, for the market, the bit shipment is expected at 20%. For the company, as always, we are aiming to achieve bit growth numbers that is higher than the market bit growth. For the third quarter, including the Solidigm volume, the QOQ growth will be about 10%, so higher than 10%. For the year, again including Solidigm volume, the growth will be about 70%.
Now on your second question about the inventory at the end of the second quarter. Now for both DRAM and NAND, the inventory level has gone up by around one week. As has been explained earlier, the bit shipment plan has been reduced from the initial plan. As you would also know, in the memory industry, what has already been invested in, for example, for the CapEx or the capacity. What's already been invested in then what comes out of them, for example, the production volume cannot be reduced or changed. This actually has been the cause of the memory industry's difficulties over the past few decades, and this has been the biggest cause of the memory industry's cyclicality over the past few decades.
Again, on one hand, we have the volume that is pretty much fixed, but the bit shipment plan has been reduced. This translates into higher level of inventory, and that has resulted in one week higher inventory for both DRAM and NAND. Now, having said that, we do not intend to push too strong to sell down our inventory given the difficulties in the market and also, given the higher level of inventory held by the customers as well. We do not want to have too much of an aggressive selling so that we will cause any difficulties in the market. That is not the kind of positioning that we are going to take.
For the higher level of inventory that is expected, we will also take into various factors into consideration for next year's CapEx planning.
The following question will be presented by S.K. Kim from Daiwa Capital Markets. Please go ahead with your question.
Now, first of all, congratulations on the good performance. I also have two questions, one on DRAM, another on NAND. Now, as has been explained in the presentation, the demand for data center appears to remain solid. There has been some delay in the launch of the server CPU that is going to support the demand for the DDR5. This is potentially going to affect the demand for DDR5 or in the second half or the production of the DDR5.
Regarding the DDR5 as well as other premium DRAM products, what is the company's policy down the road? The second question on NAND. Now it actually last night, Micron has announced the production of NAND 232-layer. It appears as if the competitor's ramp up is occurring at a faster speed than expected. What is the company's plan or strategy down the road?
우선 첫 번째 질문 주신 DDR5 전망과 프리미엄 제품 전략에 대해서 말씀드리겠습니다.
Now, on the first question about the DDR5 outlook as well as the strategy for premium DRAM products.
말씀하신 것처럼 올해 그 일부 CPU의 지원에 따라서 DDR5 향 demand가 사실상 축소된 것은 맞습니다. 하지만 그 내년 초부터 다시 이제 본격 ramp-up 될 걸로 현재 전망이 되고 있고, 또 다른 CPU 같은 경우에는 on track으로 시장 그 engage를 하고 있어서 결국은 한, 두 분기 정도 shift 된 거지. 큰 틀에서는 내년부터 본격적으로 DDR5 전환이 이루어질 거다라고 보고 있습니다.
On one hand, it is true that there has been some impact on the demand for DDR5 because of the delay in the launch of a certain CPU. But then we do believe that the ramp up is going to be expected in next year. Also another CPU is engaging with market on track. We believe that the plan has been shifted by 1-2 quarters at most. In the bigger picture, we believe that it remains the same, meaning that there is going to be a shift to the DDR5 in next year.
중장기 전망 관련돼서 시장 흐름 자체가 AI 머신러닝 등 고사양 컴퓨팅에 필요한 워크로드를 요구를 하고 있고, 그다음에 그러한 분야의 중장기 성장성에 대한 의심은 없기 때문에 결국은 내년부터 DDR5로의 migration은 본격적으로 진행될 걸로 보고 있고요. 그런 면에서 당사는 1α nm의 제품 경쟁력을 기반으로 본격적인 DDR5 준비 및 적극적인 시장 확대를 진행해 나갈 계획입니다.
Now over the medium to longer term, when we look at the market trend and there is increasing workload for the high specification computing, for example for AI or machine learning. We also believe that there is a growth potential for the medium to longer term for this type of higher workload and high density computing. We do believe that there is going to be a general migration to DDR5 starting next year. To respond to this trend, the company is going to prepare the DDR5 based on our competitiveness in the 1α nm and try to broaden our position in the market.
내용 중에 그 프리미엄 제품 전략에 대한 질문이 있으셨는데요. 결국은 뭐 두 가지입니다. 하나는 고용량 제품들 확대하는 부분, 두 번째는 차별적 제품 확대하는 부분, 이렇게 크게 두 가지로 볼 수 있고요. 첫 번째 관련돼서는 내년도에 1α nm 본격 확대되면서 저희가 16, 12GB LPDDR5, 그리고 16, 24GB DDR5 등 고용량 tech 제품들을 이제 본격적으로 확대하게 되고요. 그다음에 완제품 관점에서는 특히 이제 서버 분야에 고용량 제품 내에서의 저희의 지배력이 높다고 보이기 때문에 그 부분들 지속해 나갈 계획이고요. 서두에 HBM 발표가 좀 있었는데 뭐 시장 성장세는 여타 conventional DRAM들하고 비교할 수 없을 정도로 지금 굉장히 큰 폭의 성장을 하고 있고, 자사의 그 leadership 또한 매우 높다라고 지금 보고 있습니다. 그래서 그 부분 중장기적으로 지속할 roadmap을 저희가 확보하고 있고요. 더불어서 그래픽 쪽도 저 현재로서는 GDDR6 tech로 되어 있는데 이 부분들 1, 2년 내 빨리 저희가 GDDR7 쪽으로 이행하면서 leadership을 확대해 나갈 계획입니다.
Also part of the question was about our strategy for premium DRAM products. Largely I would say that there are twofold strategy. One is to increase the mix of high density products and second is to increase the mix of differentiated products. For the first part, we intend to increase the sales of 1α nm products. For example, the 12 gig or 16 gig LPDDR5 as well as the 16 gig and 24 gig DDR5. We will increase the mix of high density technologies. Then in terms of the finished products, we do believe that we have dominance in the high density areas, for example for the servers. We intend to continue to maintain our dominance in this segment.
We discussed a little bit about the HBM, and we see that in this market particularly is growing very fast, almost incomparable to the other DRAM segments. This segment is growing very fast, and our leadership in this segment is very strong. Over the medium to longer term, we have a roadmap to continue this leadership. For the graphics right now the main technology is GDDR6. Within the next one to two years, we plan to migrate to GDDR7 so as to solidify our leadership.
NAND 관련된 답변 드리겠습니다. 사실 구체적인 답변을 드리기에 앞서
Now this is the response to the question on NAND. Before responding to it, let me make an analogy. Let's say we are hiking, then when we do we see that there are differences in pace by individuals. Also even for one particular individual, there are differences in the pace or the tempo depending on the time. Even for one individual, they can travel at different speed. We see that you know, each company or each person moves at their different tempo and their different strategy. For the particular competitor that you have mentioned, it appears as if that they have a shorter tech cadence and has announced a new migration. Yes, of course, we are taking note of the development.
The question was about the company's response to the current development and as had been reiterated several times before, for SK hynix and now all the more so with Solidigm, the plan, the strategy has always been to outperform the market. For this year, now by the end of the year, we will have the mix of 176-layer in terms of wafer to be around 70% of the fab. That will also provide us with the industry best cost in terms of the fab operation. What I'm trying to say is that the company will also march at its own tempo, and we will try to achieve the kind of economy of scale that we had not seen before.
If I add one more explanation, in terms of the development, the company plans to complete the CS for 238-layer by the end of the year and begin mass production of the 238-layer next year. When we see the market trend recently, then it's not more about the technology. In other words, it's not about who develops a better technology 1-2 quarters earlier than others. I think the focus now is more on the business aspect, meaning who can deliver the latest technology in a customer friendly manner and to get the most bit growth or revenue or profitability out of it.
The following question will be presented by Nicolas Gaudois from UBS. Please go ahead with your question.
Yes. Hi, good morning. Thanks for taking my questions. First on the back of our recent research, it seems that you have actually recently reset your initial 2023 CapEx budget lower, in light of end demand uncertainty. You kind of alluded to looking at options earlier in the call. Would you actually confirm this is generally the direction? Maybe give us some initial quantification about how CapEx and specifically with further equipment spending could trend year-over-year in 2023. Secondly, you talked earlier about server memory demand. How is specifically memory procurement from hyperscale customers trending, and in particular for server DRAM? Have you seen any adjustment at this stage from U.S. hyperscale customers in particular? For the purchasing plans into H2.
How do you see the managing inventories? Thank you.
Now on your first question. Please understand that I am feeling quite cautious in responding to this question because as has been mentioned in the presentation and also given the sensitivity of this topic, I am a bit cautious. Now, since the question was asked, because the inventory level is likely to be higher than average for the market on the whole and for the memory industry on the whole and also for the customers, it is almost unavoidable that the CapEx for next year also has to be adjusted significantly. Except for the past, the immediately past 1-2 years. Looking back to pre-COVID and further to 6-7 years ago, then we see that the memory cycle has been relatively stable in that time.
From 6-7 years ago until the pre-COVID period, the demand was consistent and so was supply. We had been seeing quite a stable market cycle until then. Then COVID hits and along with that, the supply chain issues began to occur. So much so that even when there is demand in the market because of the equipment lead time, it was difficult for us to adjust the CapEx. Now, of course, during the pandemic period, it is true that there was a big boom in demand because of the rising demand for IT products. The question at that time was how to meet such demand with appropriate supply. Now I believe we are in a different phase.
Now demand is uncertain, and if there is no flexibility in supply, then this is going to create a very difficult situation for us. Fortunately, it appears that things are going to improve for both the supply chain management as well as the equipment lead time. As a result, we believe that we will be in a better position to make our moves in accordance to the changes in the market. In the second half, including the third quarter demand, the market demand remains highly uncertain. Again, it appears that inventory is going to be higher than usual. For next year's CapEx plan, we have a number, we are looking into a number of different scenarios and also, looking into the market situation. One of the scenarios includes considerable adjustment in the CapEx budget.
We will take into consideration the different scenarios and try to respond agilely as we see, as we monitor the developments of the market in the second half. The second question on the procurement outlook by the server customers. Now we have been listening to the views of the key customers lately, and it appears that there is a common trend toward higher cloud demand. We believe that over the medium to longer term, this is going to be the direction. Now for the shorter term, there has been some impact from the parts supply issues as well as uncertainties in the macroeconomic condition. Because of this, we see that there has been an impact on the demand coming from the consumer-related segments in the first half.
This has also affected the server side. Because of this, we see that the server customers has been conservative in their stance for some time. We believe that this is going to be, this is a tactical move, in order to adjust their inventory level. Now with the top tier customers, we are continuing with our discussion with them for their planning in the second half, with the assumption that the plan for the year will be maintained. Starting this quarter, I believe that our discussion would also include planning for next year as well. There is going to be, implementation or execution of the plan in the second half, in linkage to the potential plan next year.
The following question will be presented by MinSeong Hwang from Samsung Securities. Please go ahead with your question.
Now, in order to improve profitability then it appears as if the cost is going up. In terms of the cost reduction, how much has been the cost reduction YOY?
A related question is now the company also mentioned that various options are being considered when it comes to the inventory. In order to meet the low teens % growth target for DRAM, the company also mentioned that in the third quarter it is likely to be flat, meaning that in order to meet this low teens % growth, the company will need to have significant growth in the fourth quarter, which appears to be too optimistic at this time. Also considering the equipment lead time, if the company were to make their decisions about the inventory level after monitoring and waiting and seeing things in the second half, perhaps it is going to be too late to make that adjustment next year.
What does the company think about making significant adjustment to the inventory at this time? Another question is, in the presentation, it was mentioned that the company will focus on the core business competitiveness in the year. Does this mean there has been any change to the company's strategy from previous year, which had been focused more on profitability aspect?
Well, there were so many questions in there, I do not know where to begin. Now first of all, regarding the comment that if the bit growth for the third quarter is going to be flat, then wouldn't it be difficult to reach the target of low teens% for the year for DRAM?
Well, it is true that it is going to be a challenge, but then for the fourth quarter, we are expecting the kind of demand that we normally expect at the year-end. We do believe that we will be able to achieve a very low-teens % bit shipment growth for the year. For both the company and the market. Another part of your question was the cost. For DRAM, except for 2019 when the ASP drop was particularly significant. Except for that one year, in the past 3 years, our cost reduction has been to the extent that has been more than enough to offset the drop in the ASP. Likewise for NAND or even more so.
Again, except for 2019, looking at the past three years then the ASP would sometimes go up by single digit, drop by, like, two digits. Still, our cost reduction in NAND has been enough to more than offset the ASP drop, more so than DRAM. The question is, will this remain so in the future? I would say that depends on the market circumstances as well as the company's efforts. The second part of your question was about our business operation strategy and whether there has been a change in our business focus. In other words, whether we are going to shift away from focus on profitability, more towards strengthening competitiveness. I believe that you probably meant more investment in R&D and so forth.
To give you the straight answer first, our focus on profitability remains unchanged. Now, having said that, I must also note that, there are slight differences between DRAM and NAND. For DRAM, again, we remain profitability focused, and that is because DRAM is the company's cash cow. It is where we can also invest into the future technologies and also even support the NAND business. It is also vitally important that we remain profitable in DRAM to keep investing into the next memory technologies. Now, for NAND, that does not mean that we are going to forsake profitability, for, you know, to increase market share in order to maintain a bit growth that is, higher than market. That is not the case.
Even for NAND, we will maintain our focus on profitability, which actually is low, lower than DRAM. Again, we will maintain profitability as we continue to increase market share. Also for the medium to longer term, we believe that by increasing the economy of scale, we will be able to reduce costs, which will then translate into profitability.
The following question will be presented by Sean Kim from Meritz. Please go ahead with your question.
My question is related to the previous question. Regarding NAND development, the company made the analogy of hiking and also emphasized the importance of maintaining our own pace. When we look at the memory environment, the memory industry, what we see is that the ASP tends to converge on the cost by of the leading companies.
What this means is that it's not only important to maintain your own pace, but we also have to be mindful of the cost reduction coming from other makers. Another question related to this is, for the acquisition of Solidigm and Solidigm in the first half on a non-GAAP basis, the operating profit margin was in the single digits. We see that of course, Solidigm has differentiated applications. I wonder what the company's plan is to improve profitability of Solidigm.
Well, before I respond to the question, let me just make a clarification, because of the internal miscommunication, this is not going to be the last question. We intend to take a few more questions.
최소한의 규모의 경제 혹은 마켓을 리드하는 규모의 경제를 달성하는 것의 중요성을 말씀드렸던 겁니다.
Now to your question, perhaps again, I was not entirely accurate in my explanation about how the company intends to remain on its own path. It does not mean that we will just, you know, be single-mindedly remain on our own path, being oblivious to other players developments. That was not the case. It's just that the point I was trying to make was about the importance of the economy of scale, because, when we look at the market, then we see that the competing on the basis of economy of scale is becoming more and more important. Now, before acquiring Solidigm, our market share was in the low 10%.
By the end of the year, we, our plan is to reach around low 20% and over the medium to longer term, of course, our target is to reach a much higher market share and along the way then it would also require trillions of R&D cost. Compared to this then what it is, what is important for us to is to reach economy of scale so that we will be able to lower the unit cost as much as possible, because that is going to be a key source of our competitiveness. That was the point that it is very important to reach economy of scale and we intend to remain on that path.
아까 customer delivery가 중요하다 말씀을 드렸는데, 물론 Hynix는 low-end 기준에서도 저희의 tech migration 속도나 혹은 새로운 선도 기술의 ramp up 측면에서 사실 이제는 마켓 선두권으로 올라섰다고 자부하고 있습니다. 다만 DRAM과 달리 NAND는 솔루션에 기반한 제품이고, 따라서 고객을 이해하고 전체적으로 시장이 어떻게 무엇을 원하는지 이해하는 것이 중요하다고 생각하고 있습니다. 그런 측면에서 Solidigm은 Intel의 one of division이었고, 특히나 서버 고객들을 대상으로 한 enterprise SSD market에서 그러한 기술적인 역량과 고객을 이해하는 측면에서 탁월하다고 생각을 하고 있습니다.
I also mentioned earlier about the importance of customer friendly product delivery. Of course speaking purely on the technology front, SK hynix has already become a market leader in terms of tech migration or ramping up leading technologies. What I was trying to say is that unlike DRAM, NAND is more of a solution based product. It is very important for us to have a good understanding of the customers, to have a good understanding of what the market actually wants. Now, Solidigm, having been a division of Intel, has been in the Enterprise SSD business for the server customers. They have very good technological capability and very good understanding of the customers.
그 측면에서 Solidigm에서, 이제는 사실은 마켓 선두기업에 사실 그 업계를 견줄 정도로 저희의 역량이 향상될 것으로 기대하고 있고, 또 한 측면에서 보면, 특히나 SSD, Solidigm은 기존에 이제 그 NAND 자체가 Intel에서 제한적이었기 때문에, 사실 고객을 많이 서브 하지는 못했지만 고객을 서브 할 수 있는 역량을 갖추고 있습니다. 저희보다도 훨씬 더 diversified 되어 있는 customer base를 갖고 있고 고객에 대한 이해를 갖고 있고요. 고객들을 timely하게 그 고객들이 원하는 제품을 공급할 수 있는 측면에서, 기존의 SK hynix의 역량을 대폭 향상시킬 것으로 생각하고 있습니다.
Through Solidigm, we believe that we will be able to further solidify our market leadership and development capability and also in the SSD market. Now, Solidigm is now in the SSD market, because Intel is limited, Solidigm was not able to serve as many customers as possible or as it could have. But it still has the ability to do so because it has the know-how of servicing customers, and they have a very diversified customer base. They also have the timely delivery capability as well. I believe that such capabilities of Solidigm greatly enhance SK hynix's capabilities in many different ways.
수익성 측면에서 마지막 말씀을 드리면, 올해 non-GAAP 기준으로는 Solidigm이 흑자를 달성할 것으로 보이고, 일회성 비용이 포함된 U.S. GAAP 기준으로는 소폭의 적자를 보일 것으로 예상을 하고 있습니다. 하지만 SK hynix, Solidigm을 포함한 SK hynix 전체로는 여전히 NAND 사업은 지난해에 이어 흑자를 달성할 것으로 보이고요. 그 과정에서 이 integration 과정이 효과적으로, 효율적으로 이루어질 수 있도록 하는 데 전사의 역량을 집중하고 있습니다.
In terms of profitability, on non-GAAP terms, Solidigm is going to be in the black this year. In terms of the U.S. GAAP, which includes one-off costs, then it is going to be slightly in the red. Overall, between SK hynix and Solidigm, the NAND business is going to be profitable for the year. In this process, we will be focusing all our capabilities to make sure that the integration process is going to be efficient and effective.
The following question will be presented by Ricky Seo from HSBC. Please go ahead with your question.
My first question is about the exchange rate and its impact on the operating profit margin. In the second quarter, it seems as if the won on average depreciated by about 5%. What has been the impact on the company's operating profit margin? And also about the structure of the exchange rate affecting the company's operating profit margin. Could you explain a bit more about this? For example, for every 100 won change in the exchange rate to the dollar, what would be the impact on the company's OPM? And the second question is, the company's inventory level was mentioned earlier, but then also what is the company's understanding of the inventory level by the customers, by application or by channel?
Regarding your question about the exchange rate going up, in other words, the Korean won depreciating and its impact.
Now, in the second quarter, it is true that there has been a 5% rise in the Korean won's exchange rate to the dollar, meaning that it has depreciated by about 5%. For the company, we receive our payments 100% in U.S. dollars. That has translated to roughly KRW 500 billion increase in our sales so far. Now of course there is the other side, meaning the cost side, the expenses where we also pay in foreign currency, and that is about 40% of the total cost. If we exclude the yen, against which the Korean won appreciated, so that is about 5% out of the 40%.
That means that there was impact on about 35% of our cost coming from the exchange rate rising or the Korean won depreciating. If we are to exclude that, then that means that the net impact of the Korean won depreciating against the dollar is about KRW 400 billion. To your question about the impact on the company's P&L for every 101 difference in the exchange rate to the dollar. Of course, it is difficult to give a definitive answer to that. I think the numbers from this quarter gave a lot of hint. We can look at the numbers for this quarter and also the impact of the exchange rate on the company's revenue as well as the expense side, as well as the net impact.
For the revenue it was over KRW 500 billion increase and for the cost side it was about KRW 100 billion. For the net impact for the operating profit it is about KRW 400 billion. I guess, we could use these numbers to make quarterly simulations. About the inventory per customer or per application. Now, the assessment of the customers or other's inventory is, you know, cannot be accurate from our side. Also depending on the source of information or depending on the viewpoint, the information that we get, tends to vastly differ. Now having said that, we can still look at the overall trend, and the trend is pointing upwards for inventory across all the applications. Of course, by product, differences exist.
What we are planning to do is to have a strategic operation of our inventory, against this backdrop. We will take one last question.
The last question will be presented by Simon Woo from Bank of America. Please go ahead with your question.
My first question is on CapEx and second is about the lower bit growth that is going to appear to be the norm compared to the previous cycles. Looking at the CapEx, it appears as if there is very little room for the company to reduce the CapEx in the second half of the year, given that this is already the end of July. Also about a half of the CapEx has already been committed to non-WFE areas. For example, you need to purchase a site and also you need to build the LNG related to the utilities. It appears that for this year, reducing CapEx is not going to happen. Looking into next year, I think it is a similar picture.
Of course, you could decide to reduce the investment in infrastructure. When it comes to equipment, given the fact that the company has to keep competing against other makers like TSMC, Samsung and also US and Japanese companies, for the company to reduce or not use lithography equipment or also decide not to procure etching equipment. I believe that is going to be highly risky, especially if demand turns out to be higher than expected. Also the competitors might decide to get the equipment that the company decides not to use in order to ramp up their production or upgrade their facilities. My question is then, does the company believe that it is possible or realistic to revise down CapEx next year either?
Now regarding the CapEx, it is true that for the past few years it has been very difficult for the company to be flexible about the CapEx spending. That is why we have repeatedly emphasized for the past few quarters that our business planning window is going to be pulled up to October from the usual year-end. Now I also mentioned that we will make our decision about next year's CapEx by looking at the developments in the market in the second half. That does not mean that we will be making the adjustment in October or November. Because for the memory side, we can pretty much forecast the next three months because the customers commit for that long.
We could also stretch that out to about 6 months, meaning that we can make the forecast for the next 6 months. When we say that we will be looking into the business plan for CapEx next year in October this year, it does not mean that we will be making that decision at that time. For the CapEx, we will be able to make that decision much earlier, sometime in August or September. Also, it is true that the company's investment in infrastructure has gone up. We do have a higher mix of infrastructure investment out of our total CapEx because it has, the sheer number has also gone up in recent years. Now, one thing that you have mentioned about the power generator, our own power generator.
Now looking at the overall power situation as well as the cost of electricity, we do believe that this infrastructure investment is worth the investment. Because over the medium to longer term it is the type of investment that will enhance the company's competitiveness and will end up significantly reducing our OpEx. Although this requires some investment, this is one investment that we intend to continue. Also as you have pointed out, we also have plans to purchase sites, for example, the Yongin site and also other land and also make investment into infrastructure. Still, I would say that the dominant part of our CapEx is equipment. There is still room for us to reduce investment in equipment.
When we look at the other players, so for both our competitors and foundries, I believe currently the consensus is not to be overly aggressive in CapEx investment, so much so that they will try to pull up the volume that is going to be needed later to make the investment now. I believe that is currently the consensus. This is not a direct response to your question, but let me also say this as a way of closing comment. Now, today, when we were, you know, doing this earnings release, we did mention that in Korean won terms, the company has achieved a historic high in business performance.
Given the less than rosy picture in the second half as well as the uncertainties for next year, I believe that the mood today was not too celebratory and perhaps it was a bit gloomier than we had hoped for. Now, in hindsight, when we look at the past decade or so, then we see that the memory industry has developed in a fairly stable manner. Of course, it is true that we had not been able to exercise flexibility in supply for some time, so that we are not able to respond agilely to changes in the demand. Also when we look at the ASP then for the past two to three quarters. Let's say two to three quarters ago, we were talking about upturn, and now we talk about downturn.
It does appear as if we are going back and forth between hot and cold too often. Having said that, if we can recover supply flexibility, then I believe that the market will also begin to normalize, meaning that we will also be able to regain relatively stable growth in the memory industry as well.
With that, we conclude the second quarter earnings release conference call by SK hynix. Thank you very much for your participation.