Cisco Earnings Show AI Networking Demand Extending Beyond Compute into Switches, Optics

Cisco's recent earnings beat came as a surprise to some investors who viewed the company as a traditional networking play, exposed to cyclical enterprise refresh patterns. However, the guidanceCompany-issued forecasts of future financial performance. and CEO commentary revealed that AI data center buildout is driving outsized demand for Cisco's switching, optical, and routing gear. Unlike Nvidia, which supplies GPUs at the core of AI clusters, Cisco supplies the networking "glue" that connects clusters together and links them to the broader data center ecosystem.

The implication is significant: data center operators are not just buying GPUs; they are building entirely new networking architectures to handle AI model training and inference workloads. Traditional Ethernet switches, storage, and optics are all being refreshed at accelerated cadences. Cisco is a primary beneficiary because it dominates enterprise networking and has strong relationships with hyperscalers who are driving AI buildout.

Mercedes Benz CEO and other tech leaders have emphasized in interviews that the bottleneck is no longer compute but infrastructure broadly: memory, power delivery, cooling, and networking all need to scale in tandem. Cisco's earnings affirm this; the company is guiding to sustained strength in switching and optics as hyperscalers build out AI colocation and interconnect capacity.

For investors, this validates the thesis that Nvidia's rally is not a one-stock story. Broadcom (AVGO), which supplies optical and switching ASICs, has also benefited. AMD, Intel, and others in the chip supply chain are all gaining as the AI capex cycle widens. However, Cisco's beat also raises a question: why has networking not taken a larger share of the AI capex narrative? The answer may be that GPUs are visible and easy to track, while networking infrastructure is more invisible but equally critical.