Cisco's chief executive said the company expects to land multiple AI-related design wins over the next six months, as orders from hyperscale cloud providers reach a $9 billion annual run-rate. The comments signal that the networking giant is carving out a bigger role in the data centers that power large-scale AI workloads.
The $9 Billion Run-Rate
Hyperscaler orders — the big cloud operators that build and run massive data centers — have climbed to a $9 billion run-rate, according to the company. That figure represents the annualized value of current order flow, not a one-time spike. It suggests sustained demand for Cisco's networking equipment as those operators expand capacity to handle AI training and inference.
The run-rate is a key metric for investors because it shows how much recurring business Cisco can expect from its largest customers. The company has been pushing into AI infrastructure, betting that its switches and routers will be essential for moving data between the thousands of chips inside modern AI clusters.
Design Wins on the Horizon
The CEO's expectation of multiple design wins in the next half-year points to a pipeline of deals that are still in the engineering and validation phase. A design win means a customer has chosen Cisco's hardware for a specific project, often after a competitive bake-off. These wins don't always translate into immediate revenue, but they lock in future orders and give Cisco a foothold in next-generation data center architectures.
Winning these deals matters because AI workloads place unusual demands on networks. Training a large language model requires shuttling terabytes of data between GPUs at speeds that can bottleneck older infrastructure. Cisco's recent advancements in AI infrastructure are aimed at easing those bottlenecks, and the company says better connectivity could accelerate AI development and deployment.
Faster AI Through Better Connectivity
The logic is straightforward: if data moves faster between compute nodes, training times shrink and inference becomes more responsive. Cisco's gear sits at the center of that data movement, and the company has been updating its product line to handle the bursty, high-volume traffic patterns that AI models generate.
That positioning is what the CEO is betting on. The next six months will show whether those design wins actually land and how much of the $9 billion run-rate they represent. For now, the company is signaling that the AI buildout is far from over — and that Cisco intends to be a bigger part of it.
