As data center construction methods are being restructured around 'speed' and 'integration' in response to the growing demand for artificial intelligence (AI) infrastructure, Supermicro has put forward a strategy of providing the entire infrastructure as a package. A key feature is the emphasis on shortening the AI data center construction period by bundling the entire process—from design to construction and operation—into a single system.

Supermicro held an online press conference in advance of 'Computex 2026' on the 29th and unveiled its AI data center strategy and key technology directions.

In the presentation that day, CEO Charles Liang presented changes in data center construction methods as a key task. They explained their core strategy as the 'Data Center Building Blocks Solution (DCBBS),' which provides integrated servers, storage, network, power, cooling, and software, rather than the existing method of procuring individual equipment.

CEO Liang stated, “We can support the entire data center construction process through a single supply chain,” adding, “It is a structure where customers can secure the elements necessary for building AI data centers through a single channel.”

This method focused on shortening the construction period by integrating the design, procurement, installation, and operation stages. The company explained that based on this, it has established a supply system expanded not only to server units but also to rack and data center units.

Supermicro emphasized 'time reduction' as a competitive factor for AI infrastructure.

According to the announcement, key competitive factors in the data center construction process include the speed of applying new technologies (time to market), the speed of product supply (time to delivery), and the time to actual operation (time to deployment).

The company stated that through pre-design and modularization, on-site implementation time at customer sites can be reduced from the existing several months to 'a few days or weeks.'

In addition, they presented a strategy to accelerate the data center's launch by applying a 'plug and play' structure that allows for immediate operation after installation.

CEO Liang explained, “Data centers must be able to go online quickly,” adding, “It is important to support customers so that they can start operations fast after their investment.”

Supermicro announced that it is expanding its production bases to meet increasing demand, including expanding the size of its global production facilities from approximately 2 million to 7 million square feet over five years, expanding new facilities in Silicon Valley, USA, operating production bases in Taiwan, Malaysia, and Europe (Netherlands), and establishing a distributed production system to respond to customers in each region.

The company explained that through this expansion of its production base, it has secured a system capable of handling orders ranging from small to large quantities.

In addition, they stated that they are responding to supply chain variables, such as memory supply issues, by coordinating schedules with customers.

As AI computational density increases, cooling and power have also been presented as key challenges. The company presented directions for improving power efficiency along with the expansion of liquid cooling technology.

According to the announcement, the supply of liquid-cooled servers is on the rise, and improving quality stability and maintenance efficiency was cited as a major challenge.

In addition, the review of Small Modular Reactor (SMR)-based power supply technology was mentioned as a solution to address data center power issues. This technology was described as a measure to supplement data center power supply.

A outlook for the AI market was also presented. The company explained that demand is expanding from machine learning and large-scale AI to agentic AI and industry-specific AI.

It was stated that, in particular, there is an increasing trend of building proprietary AI infrastructure not only at the cloud company level but also at the national and corporate levels, and that demand for data center construction continues across various industries.

CEO Liang said, “The demand for AI is spreading across various industries and regions,” adding that “the trend of companies building their own AI models is continuing.”

In terms of environmental response, energy efficiency improvement and eco-friendly design directions were emphasized.

The company explained that it is continuously pursuing low-power design and energy efficiency improvements, aiming to minimize environmental impact.

CEO Liang stated, “AI infrastructure must consider not only performance but also environmental impact,” adding, “We are pursuing a direction that increases efficiency while reducing energy consumption.”