On May 28, both companies, SK Hynix and Micron, broke the $1 trillion market cap barrier, marking a pivotal moment in the semiconductor industry as they joined an exclusive group of tech giants whose stocks are now closely intertwined with the unfolding AI revolution. The two milestones highlight the pivotal role of high-bandwidth memory chips in driving the world’s most powerful AI systems, making them the new oil of the digital economy.
That was driven in part by new investor capital flowing into memory makers as demand for HBM chips, the specialised stacked DRAM technology needed for training and running large language models, continues to surge. South Korea’s second-largest chipmaker, SK Hynix, rallied over 7% in early trading, becoming the first company in South Korea to break the symbolic trillion-dollar valuation mark. It was soon followed by the Idaho-based American giant, Micron, whose stock rose 8% on high trading volume. When combined, the two companies are worth more than many national economies, demonstrating the explosive growth in AI infrastructure spending.
The HBM Gold Rush Driving Record Valuations
At the core of this valuation explosion is HBM, an emerging memory technology that stacks multiple DRAM dies in a column to achieve much higher bandwidth than traditional chips, along with lower power consumption. For the top AI accelerator vendors, HBM is the indispensable component of data centres powering the latest generative AI workloads, delivering blistering speeds of hours, even for massive amounts of data.
Industry sources say that despite the fast pace of production, there is still a shortage of HBM supplies. As is known, SK Hynix is the current market leader in HBM technology and has been operating near capacity for months, with its new HBM3E and next-generation HBM4 sold out until 2027. It has been able to catch up fast, announcing its own advanced HBM stacks with major hyperscale customers. The outcome has been a virtuous circle: record revenue enables further revenue growth, which boosts demand from AI developers, leading to more capacity and innovation.
The analysts estimate the HBM market will reach as high as $50 billion a year by 2028, from almost nothing three years ago. The memory business is viewed as cyclical and low-margin, but that’s all changed with this hyper-growth. For the first time, memory experts can earn as much as pure-play logic chip designers.
The Ripple Effects through the AI Supply Chain
The trillions of dollars are not just about SK Hynix’s and Micron’s bottom line. They represent a new shift in the semiconductor landscape, in which memory is alongside processors as a critical gatekeeper of AI innovation. No other cloud vendors or AI startups have been able to quickly strike long-term supply deals, even bidding up the price to ensure months, and sometimes years, of access.
This boom is also changing investments all throughout the ecosystem. Advanced packaging equipment makers are reporting order books filled through 2027, and wafer fabs are reporting an order backlog extending into 2027. And materials makers are reporting that the specialised substrates and chemicals used to make HBMs are getting them a windfall of their own. Even governments are starting to pay attention; several countries are pushing to accelerate subsidies and incentives to ensure they have their own memory manufacturing capacity for economic and technological security.
It has a more indirect but impactful effect on consumers. However, the overall contraction in the memory market is driving up component prices across the board, and buyers of smartphones and PCs are currently unlikely to notice the HBM chips in their devices. Shortages are expected to continue for some time, and may eventually trickle down to consumer electronics, but that’s not why the initial thrust is enterprise or data-centre demand, some analysts warn.
While the excitement over today’s breakthroughs in market capitalisation may be consuming their attention, each of the two giants faces challenges to overcome. To support the skyrocketing demand for HBM, billions of dollars are needed to build new HBM fabrication plants and higher-level packaging lines. SK Hynix is investing heavily in its South Korea and U.S. factories, and Micron has begun construction of a huge new factory dedicated to next-generation memory technologies in Idaho.
The geopolitical risks are further complicating factors. Tensions in trade and export restrictions continue to cause supply chain issues, and both have expanded their production bases and increased investments in foundry operators worldwide as a result. Energy usage is also a growing concern; the energy consumption of HBM-intensive AI clusters is high, and this consumption is driving data centre operators to seek more efficient cooling solutions and greener energy sources.
However, the future seems very positive. The demand for faster, denser and more power-efficient memory will only increase as AI models become bigger and more complex. New markets for advanced memory solutions are anticipated in the future in various autonomous applications, scientific research and edge computing.
Not only are SK Hynix and Micron moving into trillion-dollar businesses, but they’re also shifting the value chain of the technology stack. For years, the focus in the world of computing has been on the central processor and graphics processor. Today, memory is becoming the silent enabler of the intelligence revolution, powering a more-than-quiet revolution across industries and societies.
That trend seems to have caught investors’ attention, who appear to have bet heavily that it’s in the early innings. As the adoption of AI becomes more widespread among enterprises and governments, the businesses most likely to provide the memory infrastructure that will serve these systems will reap benefits for years to come. In the era of Artificial Intelligence, the memory is not just a commodity anymore; it’s a strategic crown jewel, as proved by SK Hynix and Micron.
With the two companies marking milestones in their histories, the industry is closely monitoring their growth. Established for a few consumer internet and software companies, the trillion-dollar barrier has now been crossed by hardware innovators on the AI tide. Today, for SK Hynix and Micron, it’s not the finish line but the starting gun for an even bolder chapter in the history of semiconductors, characterised by continuous creativity, strategic growth, and the insatiable demands of AI, which require ever-higher speeds.
