Samsung Electronics, the world's largest memory chip maker by revenue, reported on Thursday, July 30, that it earned KRW 89.5 trillion in operating profit for the second quarter, driven primarily by memory chips used in AI data centers. According to the company's earnings report, consolidated revenue rose to KRW 171.5 trillion, a 28% increase from the previous quarter.

The results arrive within a broader industry pattern. SK hynix and Micron Technology both reported record quarters within days of each other. Together, Samsung, SK hynix, and Micron account for the overwhelming majority of global DRAM and NAND flash production, meaning their aligned record performance signals a sector-wide shift rather than a company-specific gain. All three are increasingly positioning themselves not as commodity suppliers but as critical partners in the AI hardware stack alongside GPU designers like Nvidia.

The findings also raise questions for investors. While artificial intelligence has undeniably transformed memory demand, the key question is whether prices, supply constraints, and consumer spending will continue to support current semiconductor valuations.

Memory Carries Samsung; Mobile Does Not

Samsung's financial results underscore how profoundly AI has reshaped its business model.

The Device Solutions division, which encompasses Samsung's memory business, generated KRW 127.5 trillion in revenue and KRW 89.2 trillion in operating profit—accounting for nearly all of the company's earnings. Samsung attributed the performance to strong demand for high-end products, including high-bandwidth memory (HBM), server DRAM, and enterprise SSDs used by AI cloud providers. These products command significantly higher margins than conventional memory because they are engineered to meet the extreme data-throughput requirements of AI training and inference workloads.

The mobile phone division told a different story. The Mobile eXperience and Networks unit posted KRW 33.2 trillion in sales but suffered an operating loss of KRW 0.7 trillion due to rising material costs. Shareholders still benefited, however, as earnings per share climbed to KRW 10,849—a 52% year-over-year increase—thanks to semiconductor profits offsetting the mobile division's weak performance.

On July 27, TrendForce published an earnings preview indicating that Samsung's quarterly earnings rose approximately 18 times compared to the same period last year, illustrating how rapidly AI memory has transformed the company's performance.

SK hynix and Micron Confirm the Industry Trend

Samsung's results align with a wider industry pattern.

SK hynix, the largest HBM manufacturer, reported KRW 79.3 trillion in revenue and KRW 60.5 trillion in operating profit for the quarter, achieving an operating margin of approximately 76%. Revenue surged 257% year-over-year, while operating profit soared 557%. The company also surpassed KRW 100 trillion in revenue for the first half of the year for the first time, according to its business results.

In late June, Micron reported third fiscal quarter revenue of $41.46 billion and GAAP net income of $28.24 billion. The U.S.-based memory manufacturer highlighted the expansion of its multi-year supply contracts with hyperscale customers, providing greater visibility into future AI-driven demand.

The strong execution by our teams enabled us to deliver record revenue and EPS. Memory is becoming increasingly strategic in the AI era.

— Sanjay Mehrotra, Chairman and CEO, Micron Technology

Long-term customer contracts are becoming a defining feature of the sector, representing a structural shift for an industry historically known for sharp boom-and-bust cycles. SK hynix reported that it has secured long-term supply agreements with approximately 10 major customers, ensuring profit stability while reinforcing its position in the AI server market.

HBM4 Emerges as the Next Competitive Battleground

The next battleground is HBM4, the latest high-bandwidth memory technology designed for AI accelerators. Each generational transition in HBM has historically reshuffled market share among the top three manufacturers, making the current ramp a critical juncture for competitive positioning.

Samsung began mass production of HBM4 in February. The company achieved transfer speeds of 11.7 Gbps, well above the JEDEC minimum standard of 8 Gbps. Additionally, in the second quarter of this year, Samsung confirmed that it had successfully delivered its first HBM4E technology samples to major clients.

SK hynix has also begun mass shipments of HBM4 and plans to increase output in the second half of the year. Micron, meanwhile, is advancing development of its own HBM4 products for upcoming GPUs dedicated to AI applications.

Beyond faster speeds, HBM4 doubles memory channels from 16 to 32, delivering improved bandwidth and power efficiency to meet the demands of larger, more complex AI models.

The commercial potential for HBM4 is considerable. TrendForce predicts that Samsung will earn over $1.2 billion from HBM4 sales this year as cloud providers accelerate AI facility development.

Supply Squeeze May Hit Phones and PCs

The AI-driven boost comes with challenges. Memory is a foundational component across nearly every computing device, from smartphones to data center servers, and manufacturing capacity is finite. As manufacturers prioritize AI memory production, less capacity remains for traditional chips. Samsung expects memory shortages to persist through the second half of the year despite increased production, as demand continues to outpace supply for server DRAMs, enterprise SSDs, and HBM.

According to TrendForce, DRAM contract prices are projected to rise 58% to 63% in the third quarter, while NAND flash prices could climb 70% to 75%. If these forecasts materialize, it would represent some of the steepest quarterly price increases in years, as AI servers absorb manufacturing capacity previously allocated to smartphones and PCs.

IDC analysis predicts that memory markets will remain structurally tight through 2027. Under IDC's downside scenario, global smartphone shipments could decline by 5.2% if manufacturers pass increased memory costs on to consumers.

The combined results of Samsung, SK hynix, and Micron demonstrate that the AI infrastructure surge extends beyond GPU producers like Nvidia. Memory manufacturers are among the primary beneficiaries of this boom, and their profits have become a significant barometer of global enterprise AI investment.