Shares of Samsung Electronics climbed as much as 5% on Tuesday after Nvidia CEO Jensen Huang confirmed the South Korean firm is manufacturing chips tied to Nvidia’s latest artificial intelligence push, offering a potential turning point for Samsung’s struggling foundry business.

The stock was last up about 4.3% at 196,800 won, after briefly touching 198,000 won, outperforming the broader Korean market’s 2.7% gain. The rally underpins renewed investor confidence that Samsung’s contract chipmaking unit—long overshadowed by losses and execution challenges—could finally stabilize as demand for AI infrastructure accelerates.

Speaking at Nvidia’s GTC developer conference in California, Huang said Samsung is producing chips linked to technology from AI startup Groq.

“I want to thank Samsung who manufactures the Groq LP30 chip for us and they’re cranking as hard as they can,” Huang said, adding that production is already underway and shipments are expected in the second half of the year.

Samsung also showcased Nvidia-related chips built on its 4-nanometer process, a node that has previously faced scrutiny over yield and competitiveness.

The public endorsement from Nvidia—arguably the dominant force in AI chips—carries strategic weight. It signals that Samsung’s manufacturing capabilities are meeting the performance and reliability thresholds required for next-generation AI workloads, a key concern among investors.

Foundry Unit Eyes Long-Awaited Breakeven

Samsung’s foundry division, which manufactures logic chips for external clients including Tesla and Apple, has posted billions of dollars in annual losses in recent years as it struggled to keep pace with rivals such as Taiwan’s TSMC.

Analysts say the Nvidia-linked production could help shift that trajectory. Sohn In-joon of Heungkuk Securities expects the business could reach breakeven by late next year, driven by improved utilisation rates and stronger demand for advanced-node chips tied to AI.

The logic is straightforward: AI chips require cutting-edge manufacturing processes, and securing high-profile customers like Nvidia increases both volume and credibility, two factors critical to closing the gap with industry leaders.

The development comes as global demand for AI chips continues to surge, driven by data center expansion and the need for high-performance inference and training hardware.

The AI boom offers Samsung a rare opportunity to reposition its foundry business as a strategic growth engine, rather than a loss-making segment. Unlike memory chips—where Samsung remains dominant—foundry services require sustained investment, advanced process technology, and strong customer relationships. Nvidia’s involvement could help on all three fronts.

However, rising memory chip prices—another key part of Samsung’s business—are beginning to weigh on downstream demand, particularly in smartphones. That could reduce orders for certain types of logic chips, limiting near-term gains for the foundry unit.

Sohn warned that weak mobile demand could continue to drag on earnings, even as AI-related orders improve. In addition, Samsung still faces stiff competition from established foundry leaders, meaning sustained profitability will depend on its ability to deliver consistent yields and secure long-term contracts.

Attention is also turning to potential collaborations. Advanced Micro Devices CEO Lisa Su is expected to meet Samsung Chairman Jay Y. Lee in South Korea, with discussions likely to centre on cooperation in memory and logic semiconductors. Any partnership with AMD could further strengthen Samsung’s position in the AI supply chain, particularly as chipmakers look to diversify production and reduce reliance on a single foundry provider.

Notably, the Nvidia-driven boost marks a potentially pivotal moment for Samsung. After years of heavy investment and uneven results in its foundry business, the company now has an opportunity to align itself with the fastest-growing segment of the semiconductor industry.

Analysts note that if it can convert that momentum into sustained orders and improved profitability, Samsung may finally begin to close the gap with rivals and transform its foundry unit into a core pillar of its semiconductor strategy. For now, the market reaction suggests investors are willing to give the company the benefit of the doubt—provided it can deliver on the promise of the AI era.

SK Group Chairman Chey Tae-won said Monday that the global shortage of semiconductor wafers is expected to continue until at least 2030, driven by relentless demand for high-bandwidth memory (HBM) and other advanced chips powering artificial intelligence applications.

Speaking to reporters on the sidelines of Nvidia’s GTC developer conference, Chey — whose SK Hynix is the world’s leading supplier of HBM to Nvidia — attributed the persistent bottleneck to the sheer volume of wafers required to produce HBM.

“AI actually wants to have a lot of HBM, and once you make the HBM… we have to use a lot of wafers,” he explained. “So we need some time to build up more wafers, at least four to five years. The current shortage could continue until 2030, so we expect more than a 20% shortage of the wafers.”

SK Hynix holds a dominant 57% share of the global HBM market and ranks second in overall DRAM with a 32% share, according to Counterpoint Research. The company’s position as Nvidia’s primary HBM partner has made it a critical node in the AI supply chain, but also exposed it to acute capacity constraints as hyperscalers and AI developers race to secure next-generation memory.

Chey indicated SK Hynix is actively working on strategies to stabilize DRAM prices amid the supply-demand imbalance.

“I cannot just announce right here, but I guess that our CEO is going to announce a new plan for how to stabilize the price of the DRAM,” he said, without providing further details.

The comment suggests potential production adjustments, pricing discipline, or contractual mechanisms to prevent sharp volatility that could disrupt customers and end markets.

When asked about expanding chip manufacturing in the United States — where many of SK Hynix’s largest customers (including Nvidia, AMD, and major cloud providers) are based — Chey emphasized the practical challenges. Establishing overseas plants requires secure access to power, water, construction readiness, and engineering talent — resources that cannot be mobilized quickly.

“Accordingly, he said this could not be done easily on demand,” adding that the company is currently concentrating production in Korea.

However, Chey confirmed SK Hynix is reviewing a potential U.S. American Depositary Receipt (ADR) listing to broaden its global investor base. Such a move would increase visibility among U.S. and international institutional investors, potentially improving liquidity and valuation multiples while reducing reliance on the Korean domestic market.

Chey also addressed the impact of the ongoing Middle East conflict, which has driven Brent crude above $100 per barrel in recent sessions due to disruptions in the Strait of Hormuz. He described the situation as creating “a lot of difficulties” due to elevated energy costs, prompting SK Group to actively seek alternative energy sources. While he did not specify particular options, the comment aligns with broader industry efforts to secure reliable, cost-competitive power for energy-intensive semiconductor fabs.

SK Hynix shares rose 2.7% in early Tuesday trading in Seoul, outperforming the benchmark KOSPI’s 2.4% gain and reflecting investor relief that the wafer shortage forecast, while extended, was not accompanied by any immediate negative surprises on production or customer demand.

The chairman’s remarks come amid heightened global attention to AI supply-chain bottlenecks. Nvidia CEO Jensen Huang has repeatedly warned of compute and memory constraints limiting AI model scaling, while TSMC and other foundry leaders have flagged wafer and advanced packaging shortages as key limiting factors through the late 2020s.

Chey’s timeline — a wafer shortage persisting until 2030 — aligns with industry forecasts from TSMC, Samsung, and memory analysts who point to multi-year lead times for new wafer fab equipment, cleanroom construction, and power infrastructure. The comment reinforces expectations of sustained high pricing power for HBM suppliers, particularly SK Hynix, which has outpaced Samsung and Micron in HBM3E and next-generation HBM4 development.

The potential U.S. ADR listing would follow a growing trend of Korean tech giants seeking greater international exposure: Samsung Electronics and LG Chem already have ADRs, and SK Hynix’s move could enhance its ability to fund ambitious fab expansions and R&D while diversifying its shareholder base.

With wafer capacity expected to remain tight for the remainder of the decade, SK Hynix is believed to be well-positioned to maintain its leadership in HBM — provided it can navigate geopolitical risks, energy cost pressures, and execution challenges in scaling production.