Supermicro reports Q1 FY2026 revenue decline attributable to delays in AI system deliveries.
Supermicro reported sales of $5.0 billion for the first quarter of fiscal year 2026, ended September 30, 2025, down from $5.9 billion a year earlier, primarily due to delays in AI system deliveries. Despite maintaining its position as a leading provider of AI-driven data center infrastructure, the company's deep exposure to artificial intelligence projects did not translate into growth during the quarter. Nearly $1.5 billion in revenue was deferred to the next quarter after customers requested last-minute configuration upgrades to complex GPU-based systems, which required additional integration, testing, and validation that extended build times and postponed revenue recognition.
Production bottlenecks intensified as Supermicro faced increasing complexity of next-generation GPU racks and longer lead times for key components such as advanced cooling modules and AI accelerators. The company's gross margin narrowed to 9.3 percent, reflecting higher costs tied to system integration and customization for AI deployments. While demand for AI infrastructure remains robust, these logistical and supply challenges limited the company's ability to fulfill all orders within the quarter, with AI GPU platforms accounting for more than 75 percent of first-quarter revenue.
Supermicro reported more than $13 billion in orders for its Nvidia Blackwell Ultra-based GB300 product line, with the company expecting second-quarter revenue of $10.0 billion to $11.0 billion and raising its annual forecast to $36 billion, up from $33 billion previously. CEO Charles Liang said the company is evolving into an AI and data center infrastructure provider through its Data Center Building Block Solutions (DCBBS), which streamline deployment and lower costs, with revenue of at least $36 billion expected for fiscal year 2026.
Supermicro is expanding its Data Center Building Block Solutions as a key growth driver, offering integrated systems combining servers, storage, cooling, power, and management software designed for rapid AI data center deployment. The company is ramping up global production capacity by adding facilities in the United States, Taiwan, the Netherlands, Malaysia, and the Middle East, targeting production of up to 6,000 racks per month this fiscal year, including 3,000 direct liquid-cooled racks. Supermicro is preparing for next-generation AI platforms based on Nvidia Vera Rubin and AMD Helios chips, while noting growing adoption of edge AI across manufacturing, telecom, and retail sectors.