Deployment services have emerged as the key competitive front in AI infrastructure, as buyers find that receiving hardware is only part of getting an AI factory running. With GPU supply tight and refresh cycles quickening, neoclouds and enterprises are now evaluating suppliers on how fast systems actually go live. Super Micro Computer Inc., which reported a $60 billion order backlog in July, has extended its server building-block model to cover whole data centers and is prioritizing speed over GPU count. Matthew Thauberger, Supermicro's chief revenue officer, said neoclouds want factories running quickly, while enterprises have come to view inference and agentic AI as permanent fixtures.
Thauberger said customers expect more than delivery, wanting systems online as fast as possible because time is money. He spoke with theCUBE Research's John Furrier and Gemma Allen at SC26 in an exclusive broadcast on theCUBE, SiliconANGLE Media's livestreaming studio. The discussion covered how deployment services are driving the race to time to online and how liquid cooling retrofits help older facilities handle modern AI racks.
According to Thauberger, allocation and deployment are the two biggest requests Supermicro receives. Its service teams bring systems up, run customer workloads and hand off only once everything is operational. He acknowledged the constrained supply environment is no secret, saying customers ask Supermicro how quickly it can secure allocation, bring systems online and get people on site to deploy them.
Older facilities present a separate obstacle, since many were not designed for the liquid cooling and power density of modern AI racks. Supermicro offers retrofit options including sidecars and in-row coolant distribution units, Thauberger said. The company has also added pre-assembled AI racks to its building-block lineup to shorten installations.
Thauberger said Supermicro has the largest liquid cooling capacity based in California. He said the company's U.S. headquarters and the volume of liquid cooling capacity it holds inside the United States allow it to deploy to AI factories quickly domestically.
The interview is part of SiliconANGLE's and theCUBE's coverage of SC26. TheCUBE is a paid media partner for SC26, and sponsors of theCUBE's event coverage do not have editorial control over content on theCUBE or SiliconANGLE.
More hardware news from TechManNews.




