SK Hynix, a global leader in memory semiconductors, announced a monumental second quarter, reporting an operating profit that soared past $6.6 billion, marking the highest quarterly performance in the company’s history. This record-breaking achievement was largely fueled by robust demand for AI-specific memory and the accelerated sales of its cutting-edge HBM3E 12-layer (5th-generation High Bandwidth Memory).
On July 24, the South Korean chipmaker unveiled its Q2 financials: a staggering $16.1 billion in revenue, $6.6 billion in operating profit—boasting an impressive 41% operating profit margin—and a net profit of $5.1 billion, with a 31% net profit margin. These figures not only eclipsed the company’s previous record set in the fourth quarter of last year but also surpassed market expectations, which had forecasted revenues of around $15.8 billion and an operating profit of $6.5 billion.
“Aggressive investments in AI by global big tech companies have led to a sustained surge in demand for AI memory,” an SK Hynix spokesperson stated. “Both our DRAM and NAND flash businesses recorded higher-than-anticipated shipments, contributing significantly to this landmark performance.” The company highlighted the full-scale expansion of HBM3E 12-layer sales in its DRAM segment and increased NAND sales across all applications. “Our industry-leading AI memory competitiveness, combined with a focus on profitability-driven management, has allowed us to maintain this exceptional performance trajectory,” the company added.
Financially, SK Hynix’s cash and cash equivalents surged to $12.3 billion by the end of Q2, an increase of $2.0 billion from the previous quarter. The company also improved its financial health, with a debt-to-equity ratio of 25% and a net debt-to-equity ratio of just 6%. Notably, net debt saw a substantial reduction of $3.0 billion compared to the end of Q1.
Looking ahead, SK Hynix anticipates a continued growth trajectory in memory demand for the second half of the year. The company noted that increased memory purchases by customers in Q2 led to higher finished product manufacturing and stable inventory levels. New product launches from key customers are also expected in the latter half of the year. The escalating competition among big tech firms to enhance AI model inference capabilities is projected to further drive demand for high-performance, high-capacity memory. Furthermore, SK Hynix foresees “Sovereign AI” initiatives—where nations develop independent AI systems using their own infrastructure—as a new, long-term growth engine for memory demand.
In response to these promising market conditions, SK Hynix plans to leverage its superior HBM3E product performance and mass production capabilities to nearly double its HBM shipments year-over-year, aiming for stable revenue generation. The company is also preparing for the timely supply of HBM4 to meet future customer requirements, reinforcing its top-tier industry competitiveness. Beyond HBM, SK Hynix is set to begin supplying server-용 LPDDR-based modules later this year. For AI GPU applications, the company plans to introduce 24Gb (gigabit) GDDR7 products, expanding on its current 16Gb offerings. This diversification of its AI memory portfolio is designed to solidify its leadership in the booming AI market.
For its NAND business, the company will maintain a prudent investment strategy aligned with demand, focusing on profitability. Concurrently, SK Hynix will continue product development in anticipation of future market improvements. Key initiatives include expanding sales of high-capacity enterprise SSDs (eSSD) based on QLC (Quadruple-Level Cell) technology and building a product portfolio around 321-layer NAND to enhance its market position.
“We will make some proactive investments this year to ensure a smooth supply of critical products like HBM, for which demand visibility extends into next year,” said Song Hyun-jong, President of SK Hynix. “Our goal is to evolve into a ‘full-stack AI memory provider’ by delivering the highest quality and performance products required by the AI ecosystem exactly when needed, driving both customer satisfaction and market growth simultaneously.”
