"A Market Unlike Anything Seen Before Opens... Era of Samsung and SK hynix Leading Global Tech Arrives"
Comprehensive Expert Analysis of the Semiconductor Industry Outlook
AI-Driven Surge in Demand: "Traditional Cycles Have Become Meaningless"
Market Expected to Grow to USD 1.6 Trillion by 2030
Shift Toward a Structure Led by Samsung and SK h
As the global tech industry continues to thrive thanks to advances in artificial intelligence (AI), there is a growing consensus that the traditional "three-year cycle theory" in the semiconductor sector has become virtually meaningless. Within the industry, the prevailing outlook is that growth will continue at least through 2027. As demand for semiconductors keeps rising, there are also predictions that the leadership of the global tech market will shift to memory chip makers such as Samsung Electronics and SK hynix. While experts have traditionally believed that the semiconductor industry moves in replacement cycles of two to three years—and that the current "supercycle" would peak in 2027 before declining—the recent mood has shifted dramatically.
According to the semiconductor industry on April 13, the ongoing wave of large-scale orders from major tech companies in response to AI proliferation has made it difficult to explain current market conditions using old cycle-based concepts. Against this backdrop, claims that AI investment value has been overhyped—the so-called "AI bubble theory"—are gradually fading. As a result, there is growing analysis that semiconductor demand will continue to rise beyond 2027.
The outlook across the industrial landscape has turned optimistic. Lisa Su, CEO of AMD, has predicted that the data center semiconductor market will expand to 1 trillion dollars by 2030. McKinsey & Company has offered a similar projection, forecasting that the global semiconductor market will grow to 1.6 trillion dollars by 2030. The core of this growth is expected to come from AI chips and high bandwidth memory (HBM), which are projected to account for 62% and 31% of the market, respectively.
"Samsung and SK hynix Rising from Suppliers to Market Leaders"
As AI-driven structural demand expands, there is growing sentiment that the semiconductor industry is entering a phase of long-term growth, moving away from its previous pattern of alternating booms and busts.
Ahn Ki Hyun, Executive Director at the Korea Semiconductor Industry Association, explained, "AI does not have a replacement cycle, so it is not accurate to describe this as a cycle. Unlike the PC- and mobile phone-centered markets, this is a market we are encountering for the first time." He added, "Semiconductor demand would only fall if AI demand decreases, but that does not seem likely for the time being."
Lee Jonghwan, Professor of System Semiconductor Engineering at Sangmyung University, also noted, "HBM is ordered according to customer demand, and as long as AI demand continues, the demand for semiconductors will inevitably increase. The applications for AI semiconductors, such as in military weapon systems, will only expand."
As semiconductors become increasingly important, some predict that market leadership will shift from current AI platform-centric companies such as Nvidia to memory chip makers like Samsung Electronics and SK hynix.
Yoo Hoejun, Professor of Electrical and Electronic Engineering at KAIST, stated, "We are entering an era where, instead of a handful of big tech companies running massive data centers, each company will have its own AI inference computer. As a result, the total number of devices will increase dramatically, and demand for general-purpose memory will surge." He added, "This means the leadership could shift from Nvidia to Samsung Electronics and SK hynix. Samsung and SK hynix are gradually moving from being mere suppliers to occupying dominant positions in the market."
An industry insider commented, "The fact that Jensen Huang, CEO of Nvidia, mentioned Samsung Electronics by name this year and met directly with SK hynix reflects the elevated status of memory chip makers." Recently, Samsung Electronics posted a record-high operating profit of 57 trillion won in its preliminary first-quarter results, and there are forecasts that SK hynix's operating profit will surpass 35 trillion won.
"Driven Only by AI"...Concerns About Plateauing Growth
While AI demand is driving the semiconductor market, some analysts warn that it may be risky to be overly optimistic about the future, given that the growth is heavily dependent on a single industry. The current semiconductor sector is largely influenced by the AI data center investments of major tech firms, meaning market conditions could fluctuate depending on their investment capabilities. Furthermore, rising semiconductor prices are likely to further dampen consumer demand for IT devices. As the replacement cycle for high-priced smartphones and PCs slows, the industry is becoming even more reliant on AI-driven demand.
Therefore, while the upward trend is expected to continue for now, there are views that it could reverse at any moment. Lee Changhan, former Vice President of the Korea Semiconductor Industry Association, analyzed, "By the end of next year, forecasts for AI data centers are expected to decline. As a result, the growth rate could fall to within 10% by the second quarter of 2027."
Hot Picks Today
"When Should I Sell My $30 Million in Shares?"... SpaceX Employees Face a Happy Dilemma
- Nearly 500 Patients in a Day Bring Medical System to Brink of Collapse... Nigeria Faces Cholera Outbreak Emergency
- [Breaking] Trump: "May Order New Airstrikes in Response to Iran Dragging Out Negotiations"
- "Japan Outraged Over Korea's 'Theft'; New Variety Rights Protection Agency to Launch in August"
- Travelers Risk Being Stranded... Japan on High Alert Amid Worst-Ever Typhoon Scenario
There are also predictions that semiconductor growth could slow down once the industry transitions from the "AI learning" phase to the "AI inference" phase after 2030. Kim Yongseok, Distinguished Professor at Gachon University’s College of Semiconductor, said, "By around 2030, the AI inference market is expected to surpass the AI learning market, but it is unclear whether the current supercycle will continue beyond that. In the inference phase, general-purpose memory will be used more than HBM, so it may become difficult to realize the kind of extraordinary profits we are seeing now."
© The Asia Business Daily(www.asiae.co.kr). All rights reserved.