Cisco Locks In $5.3B of Hyperscaler AI Orders, Cuts 4,000 Jobs, and Stock Jumps 16%
Cisco doubled its AI hyperscaler order forecast to $9B, raised full-year revenue guidance, and announced a $1B restructuring to redirect investment toward silicon, optics, and security as the AI capex boom keeps reshaping incumbents.
Cisco delivered the kind of quarter that erases a decade of cautious narrative. After the bell on May 13, the networking giant reported fiscal Q3 revenue of $15.84 billion against a $15.56 billion analyst estimate and adjusted EPS of $1.06 versus $1.04 expected, sending shares up more than 16 percent in extended trading — the biggest single-day jump since the dot-com era — and lifting the stock's year-to-date gain to roughly 32 percent. The Nasdaq Composite rode the move to a fresh all-time high of 26,635 on Thursday.
The headline number was AI infrastructure orders from hyperscalers, which have now reached $5.3 billion in the fiscal year to date. Cisco raised its full-year hyperscaler AI order forecast to $9 billion, up from a prior $5 billion, and finance chief Mark Patterson told analysts that another $6 billion of revenue on the AI hyperscale side in fiscal 2027 is reasonable to expect. Cisco also lifted its full-year FY2026 revenue guidance to $62.8–$63 billion from $61.2–$61.7 billion, with networking orders growing more than 50 percent year over year and data-center switching orders up more than 40 percent.
To match the new mix, Cisco is cutting fewer than 4,000 jobs — less than 5 percent of its roughly 86,200-person workforce — and will take up to $1 billion in restructuring charges, $450 million of it in Q4 and the rest in fiscal 2027. CEO Chuck Robbins framed the move bluntly: the companies that will win in the AI era will be those with focus, urgency, and the discipline to continuously shift investment toward the areas where demand and long-term value creation are strongest. The freed-up dollars are being redirected into silicon, optics, security, and internal AI tooling for employees.
The bullish read is that Cisco has finally cracked the hyperscaler buying cycle that long favored Arista and white-box switches. The company is now selling AI-pod-scale Ethernet fabrics, optics, and Splunk-powered observability into the same buildouts that Meta, Microsoft, Google, and OpenAI have been guiding to hundreds of billions of dollars combined this year. With Meta alone targeting $125–$145 billion in 2026 capex, the networking and optics bill of materials inside each new gigawatt-scale campus has become a multi-billion-dollar standalone market.
The risk Robbins did not dwell on is concentration. A handful of hyperscalers now drive the upside in Cisco's order book and in much of the broader semiconductor supply chain, which means a single capex pause from any of them would land harder than past enterprise cycles. For now, though, the market read Cisco's print as another data point that the AI build is accelerating, not plateauing — and that even legacy networking incumbents can ride the wave if they are willing to restructure to do it.
