Silicon Motion posts 127% revenue surge on AI storage demand
Silicon Motion Technology (NASDAQ: SIMO), the Taiwan-headquartered supplier of NAND flash controllers, posted net sales of $451 million in its second quarter of 2026, up 127% year-on-year and 32% sequentially. The results mark the sharpest growth in the company's public history and reflect a structural shift in where demand for storage controllers is now coming from: AI data centres, enterprise edge servers, and automotive systems, rather than the consumer devices that once anchored the business.
Gross margin reached 50.2%, up from 47.1% in Q1, while GAAP operating margin expanded to 22.4%. On a non-GAAP basis, operating margin came in at 23.1%. Net income (GAAP) rose to $136 million for the quarter, against $67 million in Q1 and $16 million in the equivalent quarter a year earlier.
The product mix telling the bigger story
The headline numbers are striking, but the product-level breakdown reveals the deeper strategic pivot. Sales of Ferri and Boot Drive solutions, Silicon Motion's offering targeting automotive and enterprise boot applications, surged an estimated 1,690% to 1,695% year-on-year. Embedded eMMC and UFS controller sales, which serve smartphones, IoT devices, and edge computing modules, grew 95% to 100% year-on-year. SSD controller sales for servers and client PCs rose 50% to 55% over the same period.
President and CEO Wallace Kou framed the company's trajectory explicitly: "Our shift from a consumer-focused NAND flash controller maker to a diversified leader in controllers and storage solutions, from AI infrastructure to the edge, is accelerating rapidly."
Management guided Q3 revenue at $519 million to $541 million, representing 15% to 20% further sequential growth. The company says it expects full-year revenue to more than double year-on-year, which would represent the highest annual revenue in its history. Non-GAAP operating margin is guided to 27.5% to 28.5% for Q3, indicating continued operating leverage as the enterprise and AI product lines scale. These are forward-looking estimates and carry execution risk.
Convergence read-across: storage as AI infrastructure
The Silicon Motion results are a useful gauge for a part of the AI infrastructure stack that receives less attention than GPU clusters or networking fabric. Every AI training run and inference workload requires fast, reliable storage at scale. As hyperscalers and cloud providers continue to expand their AI server deployments, demand for high-performance SSD controllers, particularly PCIe 5 and the PCIe 6 generation now in development, is rising in lockstep.
The dramatic uplift in Boot Drive and automotive storage solutions also connects to a second convergence: physical AI. Autonomous vehicles, robotics platforms, and industrial edge devices all require embedded storage that can tolerate harsh operating conditions. Silicon Motion's Ferri product line is positioned squarely in this space, and its near-vertical growth trajectory suggests that automotive and industrial OEM procurement cycles are now accelerating alongside data-centre demand.
For cross-sector investors, the story here is less about a single chip company's quarterly beat and more about what the numbers imply for the broader AI-infrastructure supply chain. Silicon Motion is a merchant supplier, meaning its customers span most of the major NAND flash vendors and leading OEMs globally. A 127% revenue expansion at this tier of the stack is a leading indicator of sustained capital deployment across the AI server and edge-compute build-out. Inventory on Silicon Motion's balance sheet grew to $673 million at quarter-end, up from $515 million in Q1, suggesting the company is positioning for continued strong demand through the second half of 2026. The company did draw on $59 million in bank loans during the quarter, the first such facility in recent reporting periods, which warrants monitoring as growth capex requirements evolve.