SiliconANGLE theCUBE · Nov 17, 2026
With Matthew Thauberger, Chief Revenue Officer at Supermicro
Shipping the hardware isn't the finish line, Matthew Thauberger says — getting it running is. He sums up Supermicro's whole pitch to customers in three letters: TTO, time to online.
"It's TTO — time to online. It's not just enough to deliver to the customers; they want it online as quickly as possible. Time is money." — Matthew Thauberger
Matthew told SiliconANGLE theCUBE's John Furrier and Gemma Allen at SC26 why speed, not spec sheets, is deciding who wins the AI-factory build-out.
The full segment is covered here so you can skip it.
Here are the 4 takeaways that matter.
Key Takeaways
Supermicro redirected its server-building-block speed at AI, into pre-integrated "data center building blocks"
Containers let customers roll in and plug compute straight into a facility
Neoclouds and enterprises share one constraint: a supply-constrained market
Allocation and deployment services, not just financing, are customers' top two asks
Most data centers weren't built for today's liquid cooling and power density
Supermicro's California liquid-cooling capacity is the largest based in the US, Matthew says
GPU refresh cycles now beat the old 18-month CPU cadence, but older chips keep earning
Supermicro pitches itself as a single vendor to avoid multi-vendor "fingerpointing"
1. Building Blocks for AI
Matthew frames Supermicro's 2026 as a continuation of a long-running strength, now pointed at a different target.
Supermicro transitioned its fast time-to-market server building blocks into pre-integrated "data center building blocks" for AI infrastructure
"What we've done is we've transitioned those building blocks from the servers now to the AI infrastructure. So now we have what's called data center building blocks." — Matthew Thauberger
That includes containers customers can roll in and plug into a facility, compressing the gap between ordering compute and running it.



