articles Ratings /ratings/en/research/articles/240702-credit-faq-from-bust-to-boom-how-ai-is-uplifting-the-korean-memory-makers-13153328 content esgSubNav
In This List

Credit FAQ: From Bust To Boom: How AI Is Uplifting The Korean Memory Makers


FAQ: Applying Our Integrated Analytical Approach For Climate Transition Assessments


Analytical Approach: Climate Transition Assessments


Midyear 2024 IT Forecast Update: Robust Cloud Spending Offsets Still-Cautious Enterprise Budgets


Instant Insights: Key Takeaways From Our Research

Credit FAQ: From Bust To Boom: How AI Is Uplifting The Korean Memory Makers

The memory chip market is always cyclical, but rarely this cyclical. The sector has recently pulled out of a steep, year-long downturn, into what looks like a prolonged boom. The catalyst for this reversal? Global firms' escalating spending on the systems that power AI.

Such expenditure briefly made NVIDIA Corp. the world's most valuable enterprise, and it has also uplifted all the other entities that build the components that back AI. In particular, demand for high bandwidth memory (HBM), a type of DRAM and an important component of AI, has pulled the memory sector onto a path of high growth.

Korean memory makers SK Hynix Inc. and Samsung Electronics Co. Ltd. take up the lion's share of the global HBM market. SK Hynix in particular has established a leading sales position in the latest generation of HBM chips.

The rapid reversal of conditions for Korean chipmakers has drawn inquiries from investors. We address their frequently asked questions to examine the rise of AI and its credit impact on the entities.

Frequently Asked Questions

What is our latest view on the memory market?

We believe the market will continue its upcycle trend until 2025, at least. Demand started to increase across major end-segments in late 2023. Together with the spike in AI-related spending, the memory cycle has fully reversed and should continue its upward momentum.

Chart 1


Within DRAM, AI-related demand will propel strong sales growth for high-end specialty products such as HBM and DDR5 chips. The surge in demand for these chips, which are the latest generation memory chips that enable high quantities of data processing, will extend into 2025. This constitutes the main growth driver behind DRAM. Demand for legacy chips for use in PCs, mobile devices, and servers is subdued. It is possible that this demand could recover in the second half of 2024, and the recovery would further propel DRAM growth.

Within NAND, we expect average selling prices (ASPs) to continue to recover. Capital expenditure (capex) cuts implemented by major memory makers in 2023 on the supply side, coupled with improving conditions on the demand side, should lead to more healthy supply-demand dynamics.

Moreover, rising demand for high-density enterprise solid state drives (eSSD), fueled by AI, will also likely help the NAND market. We note that NAND remains vulnerable to oversupply given the fragmented nature of the market, particularly when compared with the concentrated DRAM sector.

Chart 2a


Chart 2b


How has spiking demand for AI specialty chips affected Korean DRAM makers?

Demand for AI has been very good for chipmakers generally, and the Korean memory makers in particular. Korean entities--especially SK Hynix--are the market leaders in HBMs, an important component within graphic processor units (GPUs). GPUs are an essential chip used in AI.

With demand climbing for HBM, Korean memory makers have been raising prices for the units, thus increasing profit. HBM chips are around five times more expensive than the latest-generation graphics DRAM chips (DDR5), we believe. It is by far the most profitable segment within the DRAM space.

We expect HBMs to form a rising part of the Korean firms' DRAM output. HBM sales will likely comprise 20%-25% of total DRAM sales for the entities in 2024, and almost one-third of DRAM sales in 2025. This compares with the high-single digit percentage of turnover for that segment in 2023.

Chart 3


What is your view on SK Hynix's competitive advantage within the HBM space, and the sustainability of its global sales lead?

SK Hynix has been the sole supplier of the HBM3 chips to Nvidia since 2022, according to media reports. Additionally, we believe the company has a higher production yield than competitors for the chip. For example, SK Hynix claims close to an 80% yield on the HBM3E, which we believe is significantly higher than that of its peers. It is unlikely that its rivals will close this yield gap over next 12 months.

We believe SK Hynix will maintain its advantage over Micron Technology Inc., a key competitor, given its greater capacity to make HBMs. Capex in this segment requires significant time and investment. Samsung may find Hynix's scale advantage to be more addressable, given the former's large capital resources.

Nevertheless, SK Hynix's edge in capacity, yield rates and its standing with major HBM customers means the entity is unlikely to lose its sales lead in the next one to two years, at least.

Chart 4


Why is Samsung trailing in the HBM competition?

When SK Hynix launched the HBM chip in 2015, Samsung had also expanded its earlier-generation HBMs (HBM2, HBM2E) and had taken a large portion of the market. However, with the chips' limited growth (at the time), high cost, and low production yields, Samsung reportedly cut back on its research and development of HBMs given a perceived lack of commercial viability. SK Hynix, however, seemingly continued its development of HBMs.

This proved to be a strategic misstep for Samsung, as the company found itself trailing within the boom in AI-led demand for the latest HBM chips. Samsung reportedly has yet to make a breakthrough in supplying HBMs to Nvidia, the largest customer for such chips.

Chart 5


How do we factor in SK Hynix's strength with HBMs in our rating assessment of the entity?

We assume SK Hynix (BBB-/Stable/--) will lower its leverage below our upside trigger of 1.0x for 2024, largely driven by its profits on HBM sales. We note that the company's EBITDA may be volatile, and that its ratio of debt to EBITDA may swing widely as earnings fluctuate.

Chart 6


Chart 7


The company's capex plans pose risks. Memory companies commonly fall into the trap of spending heavily on capex during upcycles, only to be stuck with vast excess capacity and losses during downcycles. SK Hynix incurred large cash outflows in late 2022-early 2023, after it was caught out by the chip downturn, which had coincided with a period of heavy capex.

We are always mindful of the cyclicality of the memory sector, and this risks this poses for producers. However, based on our base case we do not envision the Korean memory makers experiencing a repeat of the severe downcycle of 2022-2023, for the next two years at least.

We do see potential upside for SK Hynix should the company's performance beat our current expectations. Such performance, coupled with a disciplined financial policy, may lead to sizable positive free cash flow and debt reduction.

How might U.S.-China tensions affect semiconductor entities, including the Korean memory players?

Our base case assumes no material impact from geopolitical tensions in the coming three to five years. In October 2023, both Samsung and SK Hynix secured indefinite waivers from the U.S. government to supply chip equipment to their China fabs. We believe this enables both companies to continue their China production without disruption.

However, a material expansion for both companies within their Chinese plants seems unlikely. SK Hynix has sizable production capacity in Wuxi (DRAM) and Dalian (NAND), while Samsung also has NAND production in Xi'an. We believe that Samsung and SK Hynix will mitigate their geopolitical risks by expanding outside of China.

SK Hynix, for example, has plans to expand its fab in Korea (Cheongju). Samsung may also expand production in Korea (Yongin), coupled with its ongoing expansion in the U.S. (Texas).

How do SK Hynix and Micron compare?

We rate both equally (BBB-/Stable/--), and this captures much of the firms' relative strengths and weaknesses.

We believe that the recent rise of AI is credit positive for the memory makers, with SK Hynix the biggest beneficiary. The surge in HBM demand should lead to a significantly improvement in the company's operating performance for 2024. The company has a significant scale advantage over its peer Micron in the latest generation HBM3E chips over competitor micron, while also leading in other AI related chips such as DDR5 and high-density eSSD (NAND). We believe SK Hynix has the clear advantage on business positioning.

However, Micron has the stronger financial profile. SK Hynix has a more aggressive financial record, including a capacity for high debt and leverage. The company had amassed significant levels of debt in the recent years through aggressive spending, including its acquisition of Intel's former NAND unit (Solidigm) in 2021.

Micron's financial profile is stronger than SK Hynix's, even though Micron also amassed debt during the recent downcycle. We believe that this factor compensates for the weaker business strength, and puts the Micron rating on par with SK Hynix.

Chart 8


Chart 9


Editor: Jasper Moiseiwitsch

This report does not constitute a rating action.

Primary Credit Analyst:Ji Cheong, Hong Kong +852 25333505;
Secondary Contact:JunHong Park, Hong Kong + 852 2533 3538;

No content (including ratings, credit-related analyses and data, valuations, model, software, or other application or output therefrom) or any part thereof (Content) may be modified, reverse engineered, reproduced, or distributed in any form by any means, or stored in a database or retrieval system, without the prior written permission of Standard & Poor’s Financial Services LLC or its affiliates (collectively, S&P). The Content shall not be used for any unlawful or unauthorized purposes. S&P and any third-party providers, as well as their directors, officers, shareholders, employees, or agents (collectively S&P Parties) do not guarantee the accuracy, completeness, timeliness, or availability of the Content. S&P Parties are not responsible for any errors or omissions (negligent or otherwise), regardless of the cause, for the results obtained from the use of the Content, or for the security or maintenance of any data input by the user. The Content is provided on an “as is” basis. S&P PARTIES DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, ANY WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE OR USE, FREEDOM FROM BUGS, SOFTWARE ERRORS OR DEFECTS, THAT THE CONTENT’S FUNCTIONING WILL BE UNINTERRUPTED, OR THAT THE CONTENT WILL OPERATE WITH ANY SOFTWARE OR HARDWARE CONFIGURATION. In no event shall S&P Parties be liable to any party for any direct, indirect, incidental, exemplary, compensatory, punitive, special or consequential damages, costs, expenses, legal fees, or losses (including, without limitation, lost income or lost profits and opportunity costs or losses caused by negligence) in connection with any use of the Content even if advised of the possibility of such damages.

Credit-related and other analyses, including ratings, and statements in the Content are statements of opinion as of the date they are expressed and not statements of fact. S&P’s opinions, analyses, and rating acknowledgment decisions (described below) are not recommendations to purchase, hold, or sell any securities or to make any investment decisions, and do not address the suitability of any security. S&P assumes no obligation to update the Content following publication in any form or format. The Content should not be relied on and is not a substitute for the skill, judgment, and experience of the user, its management, employees, advisors, and/or clients when making investment and other business decisions. S&P does not act as a fiduciary or an investment advisor except where registered as such. While S&P has obtained information from sources it believes to be reliable, S&P does not perform an audit and undertakes no duty of due diligence or independent verification of any information it receives. Rating-related publications may be published for a variety of reasons that are not necessarily dependent on action by rating committees, including, but not limited to, the publication of a periodic update on a credit rating and related analyses.

To the extent that regulatory authorities allow a rating agency to acknowledge in one jurisdiction a rating issued in another jurisdiction for certain regulatory purposes, S&P reserves the right to assign, withdraw, or suspend such acknowledgement at any time and in its sole discretion. S&P Parties disclaim any duty whatsoever arising out of the assignment, withdrawal, or suspension of an acknowledgment as well as any liability for any damage alleged to have been suffered on account thereof.

S&P keeps certain activities of its business units separate from each other in order to preserve the independence and objectivity of their respective activities. As a result, certain business units of S&P may have information that is not available to other S&P business units. S&P has established policies and procedures to maintain the confidentiality of certain nonpublic information received in connection with each analytical process.

S&P may receive compensation for its ratings and certain analyses, normally from issuers or underwriters of securities or from obligors. S&P reserves the right to disseminate its opinions and analyses. S&P's public ratings and analyses are made available on its Web sites, (free of charge), and (subscription), and may be distributed through other means, including via S&P publications and third-party redistributors. Additional information about our ratings fees is available at