AI Infrastructure Revenue Is Becoming the Main Story
AI infrastructure revenue refers to the income hardware and networking suppliers earn from building the compute, networking, and optical systems that power artificial intelligence workloads inside large-scale data centers over many years, beyond the initial wave of experimental deployments and pilot projects. The most telling sign that AI is moving from hype to durable business is not in app downloads or model launches, but in the order books of companies that sell the physical plumbing behind it. Cisco and Coherent are now logging record hyperscaler orders growth and resetting expectations for what a mature AI market looks like. Their latest numbers suggest that the real winners of this phase of AI are not only model providers, but the vendors enabling massive, power-hungry data center demand at the foundation layer.
Cisco Turns Hyperscaler Orders Into a Networking Supercycle
Cisco’s AI narrative is no longer a side bet; it is becoming central to its growth story. The company booked AI infrastructure orders from hyperscalers totaling 9.3 billion in its fiscal 2026, with 4 billion landing in the fourth quarter alone. That surge underpins an AI infrastructure revenue forecast of about 7.5 billion for fiscal 2027, nearly an 88% increase from the roughly 4 billion delivered in fiscal 2026. This is not happening in isolation. Cisco calls the backdrop a broader "networking supercycle," with Q4 networking product orders up 40% year over year and an eighth straight quarter of double‑digit growth. Networking equipment sales are driving the mix: product revenue grew 24%, led by 28% growth in Networking, outpacing Security, Collaboration and Observability lines. The takeaway: AI is pulling through a structural upswing in high‑end switching and routing demand, not a short‑lived spike.
Coherent Rides the Shift From Copper to Optical in AI Data Centers
If Cisco is the face of AI networking, Coherent is quickly becoming a bellwether for AI optics. The company expects fiscal first‑quarter 2027 revenue between 2.2 billion and 2.4 billion, with a midpoint of 2.3 billion, following record Q4 revenue of about 2.05 billion, up 33.8% year over year and 13.3% sequentially. Management links this to accelerating artificial intelligence data center demand and the structural shift from copper connections toward optical connectivity inside those facilities. To meet that data center demand, Coherent has sharply increased capital investment: additions to property, plant and equipment climbed to about 1.10 billion from 440.8 million a year earlier, and inventories and receivables rose as it prepared for higher business volumes. Profitability is keeping pace, with GAAP operating income jumping from 6 million to 254 million and margins expanding strongly. In other words, photonics is becoming a core pillar of AI infrastructure revenue, not a nice‑to‑have side component.
Hyperscaler Orders Growth Shows AI Demand Is Durable
The common thread between Cisco and Coherent is who is buying: hyperscalers building dense AI data centers. Cisco’s total product orders rose 35% in Q4, or 25% excluding hyperscalers, with double‑digit order growth across every geography and customer market. That spread matters. It shows that while hyperscaler orders growth is the sharpest needle‑mover for AI infrastructure revenue, broader networking equipment sales are also strengthening, supporting the idea of a long‑term supercycle rather than a single customer‑group bubble. Coherent, for its part, speaks of "exceptional customer demand" and "multiple new growth platforms beginning to ramp" as AI data center architectures transition from copper to optical connectivity. When both a networking giant and a photonics specialist are expanding manufacturing capacity and improving margins at the same time, the market is saying that AI infrastructure investment is entering a sustained buildout phase.
Conclusion: The Quiet Power Shift in the AI Value Chain
These results mark a subtle but important power shift in the AI value chain. Attention still gravitates to model leaders and application startups, but revenue momentum is concentrating in suppliers of compute‑adjacent hardware: switches, routers, optics and related systems. Cisco’s record year, with 12% full‑year revenue growth and its highest productivity metrics in three decades, coupled with Coherent’s 22.5% full‑year revenue increase and notable operating margin expansion, signals that the lasting money in AI is flowing into infrastructure. For investors and operators, the implication is clear: the AI story is no longer just about algorithms and software subscriptions. It is about who can scale the physical grid that keeps those models fed and connected, and the companies that do so are beginning to define the next era of technology growth.
