Cisco Beats Q1 Estimates and Doubles Its AI Infrastructure Forecast

Cisco's first-quarter numbers cleared the bar on every line that matters to the Street, and the company raised its fiscal 2027 outlook on the back of AI networking demand. The stock barely moved in extended trading, which tells you something. After a 32% year-to-date run, far ahead of the Nasdaq's 9% gain over the same period, investors had already priced in a good quarter. What they wanted was proof that the AI buildout is still feeding into Cisco's revenue, and this report offers a fair amount of it.

The quarter in numbers

Cisco reported non-GAAP earnings of $1.40 per share, comfortably above the $1.33 analysts had penciled in. Revenue rose 24% from a year earlier to $18.5 billion, against a consensus estimate of $18.1 billion. GAAP net income climbed to $4.6 billion from $2.9 billion a year ago, a jump that reflects both higher sales and a cost base that didn't grow as fast.

The gross margin is the number to watch. It came in at 65.6%, in line with expectations but down from 68.1% in the same quarter last year. Memory chips have been in short supply, and that has pushed up the price of the components hardware makers need for data center gear. Cisco isn't the only company absorbing that cost, and it's a pattern you'll see across networking and server vendors until supply loosens.

Guidance is where the AI story shows up

The more interesting part of the report is the forward view. Cisco now expects AI infrastructure revenue of about $9 billion for fiscal 2027, which would be more than double what it booked in fiscal 2026. That's a striking growth rate for a company of Cisco's size, and it's the clearest signal yet that its switching, routing and optics products are landing in AI data center builds, not just in the enterprise campus.

For the current quarter, Cisco guided to non-GAAP earnings of $1.34 to $1.36 per share on revenue of $18.0 billion to $18.2 billion. Analysts had been looking for $1.23 and $17.88 billion, so the guidance sits well above consensus. Chief Financial Officer Mark Patterson framed the result as the company "unlocking significant operating leverage," which he said is why earnings are growing "much faster" than revenue. In plain terms, Cisco is adding sales without adding costs at the same pace, and that math only works if the mix is shifting toward higher-value products and software.

Why the AI pivot was necessary

Cisco spent much of the generative AI boom on the sidelines while GPU vendors and cloud providers captured most of the spending. That's the context for how the company is positioning now. Rather than chase accelerators, it's selling the networking fabric that connects them, along with the optics and security layers that keep those clusters running and protected.

That timing has its risks. Investors have been nervous about a possible slowdown in data center construction, and about rising borrowing costs that could squeeze debt-funded build-outs. Those are fair concerns for any vendor whose AI revenue depends on capital spending by a relatively small group of large buyers. Cisco's guidance suggests the order books aren't showing that slowdown yet, though a forecast is still only a forecast until shipments and customer deployments confirm it.

Software and agents as the next layer

Chief Executive Chuck Robbins told analysts the company is in the "very early stages of a multiyear network supercycle," and said he's confident Cisco's AI networking technology can keep capturing a meaningful share of the opportunity. He pointed to customers turning to Cisco to "scale, secure and optimize their environments for the agentic AI era."

That phrase points to where Cisco is heading next. The company is building software for businesses that want to deploy autonomous AI agents, and it's selling security tools meant to protect those agents and the infrastructure underneath them. For a networking vendor, that's a meaningful shift. Agents that take actions across systems need tightly controlled access, and the security question becomes less about perimeter defense and more about which identities, human or automated, can do what, and for how long. Cisco is betting that the same vendor that runs the network will be trusted to govern those permissions.

Whether that bet pays off depends on adoption curves that nobody can measure precisely yet. Enterprise agent deployments are still early, and the tooling around them is changing quickly. Still, a networking company that owns the traffic path and sells the access controls has a logical position in that stack.

What to watch

Analyst reaction was measured. TECHnalysis Research's Bob O'Donnell told Reuters that a small dip in gross margin "is not a major concern given higher memory and component costs," and that the raised forecast was what the market most wanted to see. That's a reasonable read. The margin trend bears monitoring, but a strong top line and a credible AI revenue target carry more weight right now.

My read is that the next few quarters will test two things: whether the $9 billion AI infrastructure figure holds up as customers move from planning to deployment, and whether Cisco can keep gross margins from sliding further while memory costs stay high. If both hold, the company's claim that it's early in a multiyear cycle will look credible. If either slips, the stock's year-to-date gains will leave less room for error.