CSCO Q3: AI Networking Demand and Product Refresh Are Driving Real Momentum
If you build products, your roadmap depends on infrastructure trends. Cisco just posted a strong Q3, and the signal is clear: AI networking, campus refresh cycles, and cloud-first security are shaping what ships, how it's secured, and where it runs.
Here's what matters for product development leaders making decisions on capacity, security architecture, and edge deployments over the next 12-24 months.
The quick numbers
- Revenue: $14.88B vs $14.76B expected (up 7.5% year on year)
- Next-quarter revenue guide: $15.1B at the midpoint (+3% vs estimates)
- Adjusted EPS: $1 vs $0.98 expected
- Adjusted EBITDA: $5.73B (38.5% margin)
- Operating margin: 22.6% (up from 17% last year)
- Annual Recurring Revenue: $31.4B (up 5% year on year)
- Billings: $14.12B (up 9.7% year on year)
- Full-year revenue guide raised to $60.6B; full-year Adjusted EPS raised to $4.11
- Market cap: $291.5B
- Stock: $77.44, up from $73.98 pre-earnings
What moved the quarter
- AI infrastructure demand: Significant order growth from hyperscalers, with $1.3B in new orders from the same customers as last year and an expectation to at least double orders from that group in fiscal 2026. Cisco's Silicon One chips and pluggable optics are now used across major hyperscalers, signaling broader adoption of Ethernet-based AI fabrics. Learn more about Silicon One.
- Campus networking refresh: Strong demand for switching, enterprise routing, and Wi-Fi 7 as older Catalyst models hit end-of-support. Enterprises want AI-ready networks with higher throughput, better telemetry, and simpler automation.
- Security transition to cloud: Next-gen firewalls and AI-native security posted mid-teens order growth, while overall security revenue dipped due to a pivot from on-prem to Splunk cloud subscriptions (timing impact, higher long-term recurring potential). About Splunk Cloud.
- Industrial IoT and edge: Orders grew over 25%, helped by onshoring and more AI at the edge. Unified Edge-integrating compute, networking, and storage for on-site inferencing-targets use cases in retail and healthcare.
- Partnerships and regions: Expanded work with NVIDIA and G42 to support AI infrastructure and sovereign cloud needs, opening doors in new geographies and regulated environments.
Why product teams should care
AI workloads are changing network design. East-west traffic is rising, data pipelines are heavier, and teams need better observability to keep latency predictable. That affects everything from your release plans to support SLAs.
Security is moving from boxes to subscriptions. Expect budget to shift from capex to opex, with deeper integration across SIEM, SOAR, and data platforms. The timing of revenue recognition doesn't matter to users-but it impacts how vendors package and ship features you rely on.
Edge is getting practical. Real-time inferencing at stores, clinics, and factories means product teams should plan for intermittent connectivity, lightweight models, and remote management patterns that don't break under load.
Practical next steps for product development
- Run a network readiness check for inference: target low-latency paths, test east-west throughput, and confirm availability of 100G/400G optics where needed. Prioritize Wi-Fi 7 for high-density sites.
- Map Catalyst/routing refresh to your release calendar: align EOL/EOS windows with critical launches to avoid surprises, and sync with security on zero-trust and segmentation plans.
- Pilot edge AI: pick 2-3 use cases (loss prevention, patient throughput, predictive maintenance). Define on-site model update cadence, telemetry standards, and rollback procedures.
- Prepare for cloud-delivered security: model subscription costs, procurement workflows, data residency constraints, and incident response runbooks across hybrid environments.
- Validate AI fabric choices: if you're building or integrating LLM-heavy services, compare Ethernet (Silicon One) vs alternatives in terms of latency, reliability, and skill availability. Standardize on tooling for packet capture and observability early.
- Mitigate supply risk: given rising orders, secure lead times for critical networking SKUs and predefine multi-vendor substitutes.
What to watch next
- AI infrastructure order growth among hyperscalers and sovereign cloud providers
- Adoption of next-gen campus gear (Catalyst 9K, Wi-Fi 7) and edge platforms like Unified Edge
- Security revenue stabilization as Splunk cloud subscriptions scale
- Any signs of supply tightness that could affect rollout schedules
Bottom line
Cisco is executing on AI-centric networking and a broad product refresh. For product leaders, this points to faster backbones, software-first operations, and practical edge deployments-plus a security model anchored in subscriptions rather than appliances.
If your team is leveling up skills for AI-heavy roadmaps, you may find curated learning paths useful: AI courses by job role.
Your membership also unlocks:
