SK海力士2025年营收97.1万亿韩元创新高，净利润翻倍增长

Core Insights - SK Hynix reported record-breaking financial results for the fiscal year 2025, with revenue exceeding 97.1467 trillion KRW (approximately 467.47 billion RMB), a year-on-year increase of 47% [4] - Operating profit reached 47.2063 trillion KRW (approximately 227.16 billion RMB), marking a 101% increase year-on-year, with an operating margin of 49% [4] - Net profit was 42.9479 trillion KRW (approximately 206.67 billion RMB), reflecting a 117% year-on-year growth [4] Financial Performance - For the fourth quarter of fiscal year 2025, revenue was 32.8267 trillion KRW (approximately 157.96 billion RMB), a quarter-on-quarter increase of 34% [6] - Operating profit for the fourth quarter was 19.1696 trillion KRW (approximately 922.44 billion RMB), up 68% from the previous quarter, with an operating margin of 58% [6] - Net profit for the fourth quarter was 15.2460 trillion KRW (approximately 73.36 billion RMB), a 21% increase quarter-on-quarter [6] Business Segments - In the DRAM sector, sales of High Bandwidth Memory (HBM) more than doubled year-on-year, driving record performance [3] - The company has commenced mass production of the sixth generation 10nm-class (1c) DDR5 DRAM and developed a 256GB server DDR5 RDIMM module, the highest capacity in the industry [3] - In the NAND flash memory segment, SK Hynix achieved record annual sales by completing the development of 321-layer QLC products and addressing enterprise SSD demand [3] Strategic Initiatives - The company announced a significant shareholder return policy, including an additional dividend of 1,500 KRW per share (approximately 7.2 RMB), totaling 1 trillion KRW [3] - The total dividend for fiscal year 2025 will amount to 3,000 KRW per share, with a total return scale of 2.1 trillion KRW [3] - SK Hynix plans to cancel 15.3 million treasury shares, representing 2.1% of total shares, to enhance per-share value and demonstrate a long-term commitment to shareholder value [5] Market Outlook - The company anticipates continued growth in demand for high-performance memory products, particularly as the AI market transitions from training to inference, highlighting the increasing importance of memory solutions [3]