SK Hynix became the latest company to join the exclusive $1 trillion market capitalization club on Wednesday, as relentless demand for high-bandwidth memory chips used in AI servers propelled the South Korean semiconductor giant to a record valuation and lifted the broader market to new heights.

Shares of SK Hynix soared as much as 14.9% during the session before closing up 9.3%, pushing the company’s market value to 1,680 trillion won ($1.12 trillion). The milestone comes just weeks after domestic rival Samsung Electronics crossed the $1 trillion threshold on May 6, and a day after U.S.-listed Micron Technology achieved the same feat.

With Taiwan’s TSMC already a member, South Korea has now become the first country outside the United States to boast two trillion-dollar companies, highlighting its pivotal role in the global AI supply chain.

The KOSPI benchmark index surged 2.3% to a record close of 8,229.70, after briefly climbing as much as 5.1% to an all-time high of 8,457.09. The sharp intraday gains triggered a temporary “sidecar” trading curb that halted algorithmic trading. Samsung and SK Hynix together now account for roughly half of the KOSPI’s total market capitalization.

The KOSPI has been the world’s best-performing major stock index this year, surging 95% so far in 2026 after a 76% gain in 2025 — its strongest annual performance since 1999. The rally has been almost entirely driven by the AI boom and surging demand for advanced memory chips.

AI Demand Drives Record Profits and Pricing Power

Strong global demand for high-bandwidth memory (HBM) chips, essential for training and running advanced AI models like those from Nvidia, has tightened supply and driven memory chip prices sharply higher. Prices doubled in the first quarter alone and are forecast to rise by as much as 63% in the current quarter, according to industry estimates.

This pricing power has translated into record profits for the big three memory chipmakers, SK Hynix, Samsung, and Micron — even as other parts of the semiconductor industry face cyclical pressures.

Kim Young-gun, an analyst at Mirae Asset Securities in Seoul, raised his target prices for both SK Hynix and Samsung, citing sustained supply shortages.

“We expect memory chip demand to continue exceeding supply by 2028 to keep price levels high,” he said.

He increased his SK Hynix target by 18.8% to 3.8 million won per share and Samsung’s by 14.6% to 550,000 won. SK Hynix closed at 2.243 million won, while Samsung shares rose as much as 8% before ending 2.7% higher at a record 307,000 won. The gains were helped by unionized workers voting to approve a tentative wage deal, averting a strike that had threatened global chip supplies.

UBS recently more than tripled its price target for Micron, citing “the structural changes AI has driven to the entire memory complex.”

Year-to-date, Samsung shares are up 149%, SK Hynix 215%, and Micron 245%, reflecting the massive re-rating of companies at the heart of the AI infrastructure buildout.

Retail Frenzy and Leveraged ETF Mania

The rally has been supercharged by enthusiastic retail investor participation. In recent weeks, U.S. retail investors have poured billions of dollars into a new exchange-traded fund providing leveraged exposure to Samsung and SK Hynix. On Wednesday, the first South Korean single-stock leveraged ETFs linked to the two companies surged on debut, posting double-digit gains.

Kang Jin-hyuk, an analyst at Shinhan Securities in Seoul, explained the mechanics.

“Leveraged ETF buying leads to futures buying, raising futures prices and the gap with spot prices also boosting spot purchases,” he said.

Financial investment firms were net buyers of KOSPI shares worth 1.3 trillion won, while retail investors bought 403 billion won. Foreign investors, however, remained net sellers.

The Korea Financial Investment Association’s website, which provides mandatory online courses for retail investors trading leveraged ETFs, was briefly offline on Wednesday due to overwhelming traffic.

Market breadth remained extremely narrow. Of the 918 regular shares traded on the KOSPI, only 75 advanced while 826 declined, illustrating how heavily concentrated the rally has become in a handful of AI-related names.

The Implications for South Korea

The dual trillion-dollar milestones for Samsung and SK Hynix are clear evidence of South Korea’s emergence as a critical player in the global AI supply chain. The country’s heavy reliance on a few national champions in semiconductors has paid off handsomely during the current boom, but it also concentrates economic risk. It has been noted that a slowdown in AI spending or renewed geopolitical tensions could have outsized effects on the national economy and stock market.

Nevertheless, the current environment remains strongly favorable. Global AI infrastructure investment shows no signs of slowing, and memory chip demand continues to outstrip supply. For South Korea, this represents a rare period of technological and economic tailwinds that few emerging or developed economies have been able to capture so effectively.

Qualcomm Inc. has struck a significant agreement with ByteDance Ltd., the owner of TikTok, to supply millions of custom artificial intelligence chips for data centers, marking a major milestone in the U.S. chipmaker’s efforts to expand beyond its traditional smartphone dominance into the high-growth AI infrastructure market.

Bloomberg made the report, citing people familiar with the matter. Under the deal, ByteDance will procure large volumes of Qualcomm’s application-specific integrated circuits (ASICs) to power its AI agent software and support the rapid expansion of its artificial intelligence capabilities.

The deal positions Qualcomm as one of ByteDance’s key suppliers in this critical area and provides the San Diego-based company with a high-profile, high-volume customer in China’s booming AI sector.

Qualcomm shares surged as much as 8.3% on the news, hitting a new intraday record and reflecting strong investor enthusiasm for the company’s progress in diversifying its revenue streams. The move validates CEO Cristiano Amon’s strategy to grow Qualcomm’s presence in AI computing, an area long dominated by Nvidia but increasingly contested by rivals including AMD, Broadcom, and Google.

Amon had hinted at growing momentum during the company’s last earnings call, mentioning “engagement” with several potential customers without naming them. The ByteDance partnership now provides concrete evidence of that pipeline materializing.

For Qualcomm, the deal is more than just a revenue opportunity — it represents validation of its push into custom AI silicon. The company has long been a leader in mobile processors but has faced challenges breaking into the data center and AI accelerator space, where Nvidia holds overwhelming market share. This agreement gives Qualcomm a foothold with one of the world’s most aggressive AI spenders and could open doors to additional hyperscale and cloud customers.

Manufacturing the chips through partners like Taiwan Semiconductor Manufacturing Co. (TSMC) gives Qualcomm the leverage to navigate existing U.S. export restrictions on advanced AI technology to Chinese firms, as long as the chips meet legally acceptable performance thresholds. This structure allows Qualcomm to participate in the Chinese market without running afoul of sanctions, while still capturing significant value.

Meanwhile, ByteDance’s commitment underscores its enormous ambitions in artificial intelligence. The company recently increased its AI infrastructure budget by 25% to 200 billion yuan ($29.4 billion), according to the South China Morning Post. Its Doubao chatbot, China’s most-downloaded AI application for much of last year, competes directly with global leaders like OpenAI’s ChatGPT, Anthropic’s Claude, and Google’s Gemini.

The Qualcomm chips are expected to help ByteDance accelerate its AI agent development and scale its data center operations. The deal also allows ByteDance to convert its in-house chip designs into production-ready semiconductors, leveraging Qualcomm’s expertise and manufacturing relationships.

However, the partnership highlights the shifting dynamics in the global AI chip supply chain. While Nvidia remains the undisputed leader, companies like ByteDance are actively seeking alternatives and building domestic capabilities to reduce reliance on any single supplier. For Qualcomm, this represents a strategic entry point into a market projected to grow exponentially as AI adoption spreads across industries.

The deal is also seen as part of China’s continued heavy investment in AI despite U.S. restrictions. ByteDance and other Chinese tech giants are pouring billions into data centers, models, and infrastructure, creating substantial demand for chips that can be legally supplied under current export rules.

Analysts see this as a positive development for Qualcomm’s long-term positioning. Success with ByteDance is expected to lead to further design wins and help the company build credibility in the AI infrastructure space. However, challenges remain, including intense competition, geopolitical risks, and the need to prove performance and efficiency against established players like Nvidia.

Additionally, the agreement illustrates how even strict export controls are prompting creative workarounds and accelerated innovation in the wider semiconductor industry. It also reinforces the strategic importance of partnerships between chip designers, foundries, and end customers in navigating a fragmented global market.

Overall, Qualcomm’s breakthrough with ByteDance adds momentum to its diversification story at a time when smartphone growth has matured. Deals like this are expected to become increasingly important drivers of revenue and valuation for companies seeking to expand beyond traditional markets.

Therefore, this partnership is touted to mark the beginning of a more significant chapter in Qualcomm’s evolution from mobile chip leader to broader AI infrastructure player, especially as spending continues to surge worldwide.

Cerebras CEO Andrew Feldman is pushing back against the narrative of AI as an extractive force on communities, arguing that data centers, the massive facilities powering the AI boom, can and should be built in ways that deliver tangible benefits to the towns and cities that host them, rather than imposing hidden costs.

In a recent episode of Harry Stebbing’s “20VC” podcast, first published by Business Insider, Feldman, fresh from leading his AI chip company through a blockbuster IPO, criticized the industry for poor communication and execution around data center development. He pointed to Microsoft President Brad Smith’s “Building Community-First AI Infrastructure” plan as a model worth emulating.

“These can be clean, they can make jobs, they can be good for communities,” Feldman said. “We can do this thoughtfully.”

Feldman emphasized that AI companies need to approach communities with a mindset of partnership rather than imposition. He suggested practical, low-cost ways to integrate data centers into local life.

“There’s no reason why we can’t add these to communities and have the community benefit from it. And we have to do some thinking, we have all the heavy equipment out there — build a football field for the local school, build a school, add a church or a synagogue to the community. We can be good neighbors at very, very low cost,” he said.

Data Centers as Community Assets, Not Burdens

Feldman stressed that data centers must be better stewards of local resources, with companies footing the bill rather than shifting costs onto taxpayers. He criticized past practices where firms relied on outdated financial arrangements or excessive water usage.

In an email to Business Insider, he elaborated: “In some cases, they tried to pawn off costs on the local community or use outdated financial arrangements that left the community holding the bag. And in others they were wasteful of resources. This is not cool. And none of this needs to be the case.”

One practical solution he advocated is building closed-loop cooling systems to dramatically reduce water consumption. This is particularly relevant given that, according to a Business Insider report from last June, 40% of the nation’s planned and existing data centers are located in some of the most water-stressed areas in the U.S.

By following Smith’s framework, which includes paying its own way to avoid raising local electricity prices, reducing water consumption, creating jobs, and partnering with nonprofits and universities on training programs, the industry can shift public perception from skepticism to support. Smith himself noted the historical parallels.

“Whether it was canals, railroads, the electrical grid, or the interstate highway system, each era produced its own conflicts over who bore the burdens of progress. One enduring lesson is that successful infrastructure buildouts will only progress when communities feel that the gains outweigh the costs,” he said.

Addressing AI Washing and the Real Productivity Challenge

Feldman also tackled the growing public concern over AI-driven job displacement. A March Quinnipiac University poll found that 7 out of 10 Americans believe advancements in AI will lead to fewer job opportunities. He pushed back against what he sees as “AI washing” — companies blaming layoffs on the technology when the real drivers are often post-COVID over-hiring and productivity gains that are only now being realized.

“I think to date, most of the layoffs were ‘AI-washed.’ They were because we did boneheaded hiring during COVID. It is actually because a great deal of productivity gains have occurred over the years that we’re just now harvesting,” he said.

At Cerebras, the focus is on using AI to make engineers vastly more productive, not to reduce headcount. Feldman said the company wants to hire more talent, not fewer.

“If you are an engineering organization that can’t see how to take advantage of vastly more productive engineers, I don’t think you’re long for this world. I mean, the list of things I want our engineers to do is 50 times as much as we have engineers,” he said.

This perspective reframes AI not as a job destroyer but as a multiplier of human capability — provided companies invest in training and thoughtfully integrate the technology.

Feldman’s message comes at a pivotal time. As data centers proliferate to meet the enormous computational demands of modern AI, public and regulatory scrutiny is intensifying. Communities are increasingly wary of noise, water usage, energy consumption, and limited local economic benefits. By advocating for a more community-oriented approach, Feldman is attempting to shift the conversation from fear of disruption to shared opportunity.

His stance also reflects a maturing industry awareness: the AI boom’s long-term success will depend not only on technological breakthroughs but on social license and sustainable deployment. Companies that treat data centers as extractive operations risk backlash, while those that integrate thoughtfully could build lasting goodwill and smoother expansion paths.

Analysts at Goldman Sachs have raised their year-end 2026 target for the S&P 500 to 8,000 from 7,600, becoming the latest major Wall Street firm to argue that the artificial intelligence investment boom is strong enough to keep driving U.S. equities higher despite mounting geopolitical and inflation risks.

The new forecast implies another 6.4% upside from the index’s latest close of 7,519.12 and supports the increasingly dominant view among large banks that corporate earnings, rather than monetary easing, are now the primary engine behind the market rally.

“Earnings growth has powered the entire S&P 500 return so far this year, and we expect this dynamic to continue in the coming months,” Goldman analysts wrote in a note Tuesday.

The bank also sharply increased its earnings projections for U.S. companies, forecasting S&P 500 earnings per share of $340 in 2026, representing roughly 24% annual growth, followed by another 13% increase to $385 in 2027.

The upgraded outlook highlights how rapidly Wall Street expectations have shifted around artificial intelligence and its impact on corporate profitability. Only months ago, investors were focused primarily on recession risks, elevated interest rates, and geopolitical instability stemming from the war involving the United States, Israel, and Iran. Now, many strategists increasingly believe the scale of AI-related spending is large enough to offset broader economic weakness.

Goldman said companies tied directly to AI infrastructure, particularly semiconductor firms and data center suppliers, are expected to generate roughly half of the S&P 500’s total earnings growth this year.

That reflects the extraordinary concentration now shaping U.S. equity markets.

A relatively small group of AI-linked companies, including major chipmakers, hyperscalers, and cloud infrastructure providers, has increasingly become responsible for a disproportionate share of market gains. Massive capital expenditures by technology giants on AI servers, chips, networking equipment, and energy infrastructure continue flowing through supply chains at a pace analysts say remains historically unusual.

The bank noted that semiconductor stocks “at the heart of the AI infrastructure complex” have recently been outperforming even their already rapidly rising forward earnings estimates. That observation significantly suggests investors are not simply rewarding current profits, but are increasingly pricing in expectations that AI spending could continue accelerating for years.

The optimism comes as corporate America embarks on one of the largest infrastructure buildouts in decades. Companies including Nvidia, Advanced Micro Devices, Microsoft, Amazon, and Alphabet are collectively spending hundreds of billions of dollars on AI computing infrastructure, data centers, and custom chips.

That investment wave has extended far beyond Silicon Valley, boosting demand for energy producers, utilities, networking firms, server manufacturers, memory-chip makers, and industrial suppliers.

Goldman’s call adds to a broader pattern of increasingly bullish forecasts from Wall Street firms. Last week, UBS also raised its market outlook, arguing that AI-related earnings growth could help absorb the impact of stubborn inflation, supply-chain disruptions, and elevated oil prices tied to the Iran conflict.

The bullishness persists even as several macroeconomic risks continue building beneath the surface. Oil prices remain far above pre-war levels due to disruptions around the Strait of Hormuz, keeping pressure on transportation, manufacturing, and consumer costs globally. Inflation concerns have also altered expectations for Federal Reserve policy, with traders recently increasing bets that the central bank may still raise interest rates again rather than cut them.

At the same time, economists continue warning that the broader U.S. economy outside the technology sector appears considerably weaker than headline market indexes suggest. Consumer spending growth has slowed, borrowing costs remain elevated, and several sectors tied to housing, retail, and manufacturing continue showing signs of strain. Market gains have also become increasingly concentrated among large-cap technology and AI-related stocks, raising concerns about valuation risk.

Goldman acknowledged some of those vulnerabilities, noting that weak consumer demand and higher operating costs still pose threats to earnings growth. Yet the bank argued that the scale of AI investment is currently overwhelming those concerns.

The revised forecasts suggest Wall Street increasingly views AI not simply as another technology trend, but as a structural economic transformation capable of sustaining corporate profit growth even during periods of geopolitical instability and tighter monetary policy.

That belief has become the defining force behind the current bull market. Investors are effectively betting that the race to dominate artificial intelligence infrastructure will continue generating enough revenue, productivity gains, and capital spending to support U.S. equities well into the end of the decade.