"Cyclical AI Prosperity" strengthens the narrative of AI super bull market! South Korea's 2.07 trillion US dollar semiconductor gambling is coming, and the global computational power prosperity enters a cash flow strengthening cycle.

Two super giants of storage chips in the world - the Korean storage giants Samsung Electronics and SK Hynix are making unprecedented bets on the AI super wave of unparalleled value with an investment scale of billions of dollars, marking the largest scale ever investment by these two giants in cutting-edge technology. However, the proposed capacity expansion has also raised concerns: if AI spending cools down, it may lead to a painful liquidation in the global stock market under the theme of "AI computing power trading". For global investors, the unprecedented investment frenzy around AI is undoubtedly the core driver of the strong bull market in global stocks in recent years. However, for Wall Street financial giants like Goldman Sachs Group, Inc., and JPMorgan Chase, who are bullish on the continuation of the AI super bull market, the move by Samsung and SK Hynix can be seen as a significant strengthening of the logic of "cyclical AI prosperity", further driving the bullish market trend surrounding the AI computing power infrastructure. As the boom of "cyclical AI prosperity" spreads from the North American market to every corner of the world, the windfall gains earned by Asian AI computing industry chain in recent years are beginning to provide a support system for those "hyper-scale tech giants" in the North American market (such as Meta, Amazon.com, Inc., and Alphabet Inc. Class C), indirectly supporting the unprecedented AI capital expenditure environment for large tech companies and AI application leaders such as Anthropic - the low borrowing cost environment that Trump has long coveted. The unexpected gains earned by top global AI computing infrastructure manufacturers in Asia such as SK Hynix, Samsung Electronics, Taiwan Semiconductor Manufacturing Co., Ltd. Sponsored ADR, Hon Hai, and Wistron are flowing rapidly in the global economy, reflecting the flow of funds within the AI ecosystem at a global level. The AI boom is creating unprecedented cash flow for smaller economies in Asia such as Korea and Taiwan, which is leading to a narrower version of Asian savings surplus. At the same time, these Asian AI computing industry chain leaders are evolving into key forces supporting long-term and healthy bull markets in the US and global stock markets; previously, in the late 1990s and early 2000s, Asian savings surplus had kept US benchmark borrowing costs at historic lows. In the "cyclical AI investment boom," the logic of the US stock market bull can be seen as more solid than simply "big tech companies burning money to buy GPUs" because it is forming a global loop - US AI applications and cloud companies creating massive demand and generating strong AI revenue, Asian chip/storage/server supply chain benefiting from export gains, Asian surpluses partially flowing back to US assets, thereby significantly supporting the US tech giants' stock and bond financing environment and ambitious AI capital expenditures. Samsung and SK Hynix are providing "hard evidence" on the supply side and the capital expenditure in the national industry: Korea aims to double its storage chip production capacity within five years, and Samsung and SK Hynix have pledged a total investment of 32 trillion won (about 2.07 trillion US dollars), including the construction of an 80 trillion won chip cluster in southwest Korea. The significance of the "cyclical AI investment boom" framework lies in filling in the other half of the market's previous concerns: who will pay for this round of AI computing capital expenditure, and whether AI applications can convert computing power consumption into cash flow - Anthropic expects second-quarter revenue to double to $10.9 billion, a 130% increase from the previous quarter, and achieve its first quarterly operating profit, directly weakening the pessimistic narrative that "AI only burns money and has no commercial return". In this framework, the expansion of production by Samsung and SK Hynix should not be simply understood as a "risk of overcapacity," but should be seen as a real industry bet on driving global economic prosperity through AI prosperity. From caution to rapid expansion of production: Samsung and SK Hynix double down on the AI-driven storage super cycle, as the global chip supply and demand experience a decade-long gamble After these two technology giants in Korea fully support the government's plan to promote semiconductor and large-scale AI data center construction, Korean President Lee Jae-myung praised them and bowed deeply to them. This is clearly a departure from the more restrained approach to expanding chip production in the past, which has been shaped by decades of painful cycles of prosperity and depression in the storage chip industry. According to the plan, the Korean government hopes that the country can double its storage chip production capacity within five years. Samsung Electronics and SK Hynix will also accelerate the construction of semiconductor factories within the existing Yongin semiconductor cluster, significantly shortening the construction period of 7 to 12 years and releasing additional production capacity earlier. The two largest storage chip manufacturers in the world have pledged a total investment of 32 trillion won (approximately 2.07 trillion US dollars) in the semiconductor field, including the construction of an 80 trillion won chip cluster in southwest Korea and previously announced projects. Amid the surge in global AI investments, Korea is becoming the most central force among AI investment winners, driven by the absolute dominance of Samsung Electronics and SK Hynix in the high-bandwidth memory (HBM) and broader DRAM and NAND storage chip domains, which are crucial for advanced AI GPU/AI ASIC processors dominated by NVIDIA Corporation, Alphabet Inc. Class C, and AMD. SK Hynix and Samsung Electronics stock prices have risen by 307% and 179% respectively so far this year. At the time of this announcement, from super-large AI cloud companies to the world's largest consumer electronics manufacturers like Apple Inc. (AAPL.US), these super customers are demanding more storage chips in the context of a global shortage of chip supply. However, analysts point out that it takes several years to build and ramp up chip factories, which means that most of the additional production capacity will not come until the next decade's deep stage. Morningstar analyst Jing Jie Yu said, "We believe that the pricing of storage chips will still be a function of demand and supply, and the accelerated capital expenditure over the next ten years further increases the risk of long-term overcapacity." He added that the prosperity of storage chips still depends on whether hyper-scale AI cloud companies will continue to expand at the current fast pace. If the pace remains this fast, then all the cycles of prosperity and depression will continue to be absent. Lee Jong-ho, a professor in the Department of Electrical and Computer Engineering at Seoul National University, said that these investments seem to be progressing too quickly. He said, "This is the kind of investment that could determine the future of a company." "No one knows what the situation will be like in three years. We need to respond quickly when demand is strong, but afterwards, there is uncertainty in demand, and decisions should be made with caution." The profit margins of Samsung and SK Hynix's storage business have soared, which is a relatively recent phenomenon, driven by the global AI boom and severe shortage of storage chips, with prices of all types of storage chips nearly doubling in the first quarter alone. In the past, downturns in chip prices have seriously impacted these companies, leading SK Hynix to the brink of bankruptcy in 2001, and causing both companies to record huge losses in 2023. However, in recent months, as the global AI investment wave has continued to grow, ruling party lawmakers and government officials have begun to propose moving some chip factories to the southwest region of Korea. While this region is a crucial support base for them, it has been bypassed during Korea's industrial rise. According to scenes captured by reporters, Korean President Lee Jae-myung and SK Group Chairman Choi Tae-won bowed to each other at a government briefing on three major super projects held at the Blue House in Seoul, while Samsung Electronics Chairman Lee Jae-yong watched from the side. Just two months ago, when a lawmaker inquired about the plan to build a chip factory in the southwest, SK Hynix Chairman Choi Tae-won expressed doubt. He said, "I'm not sure if semiconductor expansion is necessarily the area you should enter." Both of these Korean tech giants have not explained what prompted them to change their minds. Both Samsung and SK Hynix have stated that with the government's commitment to expedite approvals, as part of their efforts to meet the surging demand, they will now be able to accelerate the construction of the Yongin project. Nomura analyst CW Chung said, "Investing in other regions may be a way to hedge the uncertainties surrounding the Yongin semiconductor cluster." However, the analyst said he believes that these companies will learn from the numerous storage chip cycles from the past. CLSA senior analyst Sanjeev Rana said, "A downturn in the storage chip industry is clearly a risk the plan faces." But "if there are signs of overcapacity, storage chip manufacturers will still retain the flexibility to adjust investment pace". Samsung's plan envisions investing 21 trillion won in chip production by 2040, but the company points out that spending may be adjusted based on market supply and demand conditions and actual global business needs. In 2024, during a period of sustained weak demand in the global chip market, Samsung suspended construction of its P5 chip factory in Pyeongtaek, south of Seoul for nearly two years, and resumed construction at the end of last year. At a demonstration attended by President Lee on Monday, SK Hynix's Choi Tae-won said the company will monitor demand conditions when determining the scale of investment, but he added that it is still difficult to fully address the supply shortage issue. However, for the Korean government, maintaining momentum is a strategic priority. Kang Hoon-sik, chief of staff to the President of Korea, said on Monday that the government will expedite approvals and seek to complete the construction of the chip cluster before President Lee's term ends in 2030. Koh Taebong, research director at iM Securities, believes that this move is a correct bet for Korea as the world enters a transformative era driven by AI. He said, "The message conveyed by yesterday's release conference, which is destined to have far-reaching implications, is that Korea is really considering becoming one of the top three AI superpowers in the world. This is not just a slogan - the government is promising to invest real money to make it happen." From the profit outlook of Claude to the expansion of HBM/DRAM/NAND production capacity: Korea's $2.07 trillion gamble on chips is rewriting the narrative of the AI super bull market from a "burning money story" to a global cash flow loop In terms of the prospects for AI revenue, Anthropic, the company that launched a series of heavyweight AI intelligent bodies in February devastating global software stocks, has recently released a heavy profit trajectory, with strong revenue data and an unexpectedly profitable path that will accelerate the construction of "AI computing power super empires" by tech giants. Anthropic expects its revenue to double in the second quarter, achieving its first operational profit, a strong performance that highlights the shift from the narrative of "burning money" to a "cash flow cycle narrative" in the AI industry. The most significant development is that the monthly ARR of Anthropic has surged by $11 billion, equivalent to the ten-year volume of the three SaaS giants Palantir, Snowflake, and DataBricks - this is an unprecedented miracle in the history of capitalism. Anthropic also expects its second-quarter revenue to soar from $4.8 billion in the first quarter to $10.9 billion, with an operating profit of approximately $559 million, indicating that advanced AI applications are not only consuming computing power but also beginning to transform enterprise programming, intelligent body workflows, network security, and data analytics into high-value token earnings. The surge in demand for Claude and AI tools is driving Anthropic toward its first profitable quarter, and the cost of computing power per $1 of revenue has decreased from approximately 71 cents in the first quarter to about 56 cents in the second quarter, demonstrating improved scale effects and reasoning efficiency in the AI application economic model. Furthermore, a research report shows that for every $100 increase in AI-related hardware imports in the US, it may increase the GDP of major Asian suppliers by $35, accelerating the formation of an "AI-driven super surplus" in Asian economies - this also means that the AI boom is no longer just a story of valuation expansion at the application layer of OpenAI/Anthropic, but a global fund cycle of "profit expansion of US AI applications - capex of large cloud companies - hard asset export of Asian AI computing infrastructure - return of USD assets". The strong performance outlook of Anthropic, coupled with Korea's $2.07 trillion gamble on chip production capacity expansion, thoroughly strengthens the "AI super bull market" cyclical investment logic of "US AI application profit generation - Asian AI computing hardware windfall - return of USD assets - continued support for AI capital expenditure", while also reinforcing the "cash flow re-circulation cycle" surrounding the AI super bull market. Therefore, the expansion of production by Samsung Electronics and SK Hynix should not be simply understood as a "risk of overcapacity," but should be seen as a real industry bet on driving global economic prosperity through AI prosperity. If AI intelligent bodies, enterprise programming, reasoning services, data analytics, AI search, and multimodal applications continue to drive token consumption, storage demand will exhibit nonlinear growth: HBM determines GPU/ASIC performance release, DRAM supports reasoning and caching, NAND/enterprise SSD carries Beijing Vastdata Technology, vector retrievals, model contexts, and AI agents' long-term memory. As wafer fab construction and capacity ramp-up take years, the massive release of new capacity in Korea will have to wait for the next decade's deeper stage, in the short to medium term, it may instead strengthen the pricing logic of "supply unable to meet AI demands". The grand narrative of the "cyclical AI prosperity" is the only one that can link the US stock market AI bull market, Asian export surpluses, semiconductor equipment super cycle, and rising storage prices: the US is responsible for creating AI demand and software monetization, Asia is responsible for providing the physical computing power base, and capital markets are responsible for re-circulating cash flow, surplus, and valuation. From observing the flow of global capital, it currently looks more like the AI super bull market has entered its second stage. The first stage was dominated by hyper-scale training clusters driven by AI GPU and ASIC; the second stage is beginning to spread to data center power chains, HBM/DRAM/NAND, advanced packaging, liquid cooling, data center CPUs, optical communication/optical interconnection, high-performance Ethernet network infrastructure/data center DCI high-speed interconnection, as well as PC, wearable consumer electronics, humanoid Siasun Robot & Automation, autonomous driving, and other AI application endpoints, along with "new AI centers" such as SpaceX representing space AI computing infrastructure spreading across the board. Goldman Sachs Group, Inc., the Wall Street financial giant, has recently raised its target for the Korean KOSPI benchmark index several times, betting that the Korean stock market, which has surged by 100% and experienced dramatic fluctuations since the beginning of the year, will continue to rise. Goldman Sachs Group, Inc.'s target for the Korean KOSPI benchmark index has been raised from 8,000 to 12,000 points, compared to the index closing near 8,470 points on Tuesday. The core logic behind this is not macro stimulus or monetary policy factors, but the long-term prosperity cycle of storage and the continued explosion of AI memory/storage chip demand, with SK Hynix and Samsung Electronics being the significant beneficiaries of the surge in demand. In the view of Goldman Sachs Group, Inc., the global AI chain bull market is far from over, and the market's main theme has shifted from the long-term "valuation expansion of software assets driven by programming/code" since 2008 to "re-pricing of AI computing power infrastructure around a series of physical assets". Goldman Sachs Group, Inc.'s latest calculations show that the global benchmark model for AI capital expenditure is expected to increase from $765 billion annually in 2026 to $1.6 trillion annually in 2031, with a cumulative capital expenditure of approximately $7.6 trillion from 2026 to 2031. The data center power demand in the US is expected to increase from 31GW in 2025 to 66GW in 2027, directly spilling over AI computing power infrastructure investments to server CPUs, DRAM/NAND/HBM, advanced packaging, liquid cooling, power equipment, transformers, gas turbines, grid equipment, data center REITs, and engineering construction. In summary, this text outlines the significant investments and developments in the global market around storage chips, AI applications, and the interplay between technology giants, financial institutions, and government policies in driving economic growth in the digital age. It emphasizes the importance of investments in AI, storage technology, and computing infrastructure as key drivers of economic prosperity and technological innovation.