Supermicro

Lore & Background

Supermicro was founded in 1993 by Charles Liang, a Taiwanese-American electrical engineer, alongside his wife and company treasurer, Chiu-Chu Liu (Sara). Liang had previously served as president and chief design engineer of Micro Center Computer. The company turned a profit within six months. In 1996, it opened a Taiwanese manufacturing subsidiary, Ablecom, run by Charles's brothers Steve Liang and Bill Liang. A European subsidiary opened in the Netherlands in 1998, and a system integration logistics center followed there in 2010. In 2012, the company opened its Taiwan Science and Technology Park, which was expanded in 2019 and came online in 2021. Supermicro moved its corporate headquarters to the former Mercury News site in North San Jose in 2014, naming it Supermicro Green Computing Park, and added a manufacturing building in 2017. Further expansion in San Jose occurred in 2021, and in February 2025 the company began building its third California manufacturing campus, intended to increase production of liquid-cooled servers. In July 2025, Supermicro expressed interest in expanding manufacturing in Europe to meet AI demand. In October 2025, it created a subsidiary focusing on U.S. federal agencies. From November 2025 to June 2026, it added eight production facilities in the United States, including a Data Center Building Block Solutions campus in San Jose. As of June 2026, it planned two additional facilities in Taiwan and also operated a Malaysian facility.

Reader's Guide

Supermicro's significance lies in its early and sustained focus on high-performance, energy-efficient server technology. In 1995, it released one of the first dual-CPU server motherboards, followed by a four-CPU board. Energy-saving servers became a priority after 2004, influenced by founder Charles Liang watching the film The Day After Tomorrow. In 2008, Supermicro was among ten companies brought onto the Hyperion project by the National Nuclear Security Administration's Lawrence Livermore National Laboratory to develop high-performance computing testbeds for nuclear stockpile stewardship. The company's close ties to Nvidia have been central to its recent growth: in July 2024, Supermicro provided half the servers for Elon Musk's AI startup xAI, helping build a 750,000-square-foot Memphis data center for the Colossus supercomputer in 122 days. By November 2025, Supermicro had received $13 billion in orders for systems based on Nvidia's Blackwell Ultra GB300 architecture, and its revenue had grown from $7 billion to $22 billion over two years. In January 2026, it expanded liquid-cooling manufacturing capacity for Nvidia's Vera Rubin and Rubin AI platforms. At Computex in June 2026, Supermicro demonstrated a VR200 NVL72 rack using Vera CPUs and Rubin GPUs with proprietary liquid coolant, and released ARM-based rack-scale AI systems. In September 2026, it released a workstation built around Nvidia's GB300 Grace Blackwell Ultra superchip with 748GB of RAM. The company also maintains relationships with AMD, Broadcom, Intel, Samsung, and Micron.

Did You Know?

• Supermicro turned a profit within six months of its founding in 1993.
• In 2004, the company began developing energy-saving servers after founder Charles Liang watched the film The Day After Tomorrow.
• Supermicro provided half the servers for Elon Musk's AI startup xAI, completing a 750,000-square-foot data center in 122 days.