RockstarMarkets
All news
Markets · Narrative··Updated 1h ago
Part of: AI Capex

AI Infrastructure Demand Spreads Beyond NVIDIA: Cisco Soars, Broadcom Lifts

Cisco's strong Q1 earnings revealed that AI networking demand is widening into switches, optics, and scale-out infrastructure, not just GPUs. The narrative shifts from NVIDIA-centric capex to broader semiconductor and networking buildout, lifting tech stocks including AVGO and broadening the AI trade beyond single names.

R
Rocky AI · RockstarMarkets desk
Synthesised from 8 wires · 49 mentions in the last 24h
Sentiment
+65
Momentum
75
Mentions · 24h
49
Articles · 24h
45
Affected sectors
Related markets

Key facts

  • Cisco Q1 earnings showed AI networking demand widening into switches, optics, and scale-out infrastructure
  • CFOs of MSFT, GOOGL, AMZN, META, AAPL all cited memory constraints on earnings calls last month
  • Micron trading at 7x forward earnings despite memory capex tailwinds
  • US approved NVIDIA H200 chip sales to 10 Chinese companies
  • Jensen Huang spotted in Beijing; interpreted as green light for China sales acceleration

What's happening

The artificial intelligence infrastructure trade is maturing beyond the GPU-as-chokepoint story that dominated 2024 and early 2025. Cisco's earnings results highlighted a critical inflection: the capex cycle is now flowing into ancillary layers of the data-center stack, specifically networking hardware that glues together the sprawling compute clusters required for large language models and agentic systems. This shift matters because it signals that enterprise and hyperscaler AI spend is moving from proof-of-concept into full-scale production deployments that require complete architectural refresh.

On the earnings call front, this narrative has crystallized across multiple megacap earnings. In consecutive days last month, the CFOs of Microsoft, Alphabet, Amazon, Meta, and Apple all cited memory constraints as the binding constraint on AI expansion, not compute. This drove a reassessment of Micron Technology's valuation; despite the company trading at just 7x forward earnings (a steep discount to growth comps), fund managers had underweighted the memory thesis. Broadcom, the fabric-layer semiconductor provider, and Cisco, the switching and routing giant, both benefit from this reallocation, as do optics specialists that bridge data centers and the cloud infrastructure providers building out their internal networks. The bull case no longer depends on whether NVIDIA can sustain 200% growth; it depends on whether the broader ecosystem can keep pace with deployment velocity.

Geographically, this opens a window into China's AI push. The US government approving sales of NVIDIA's H200 chips to ten Chinese companies signals that even with restrictions in place, the waterfall of approved silicon is broad enough to support China's training infrastructure ambitions. NVIDIA CEO Jensen Huang was spotted in Beijing, a detail interpreted by market participants as a green light for accelerated sales cycles. If the US-China tech relationship stabilizes during the Trump-Xi summit, expect Chinese capex to become a tailwind for the entire AI semiconductor and networking complex, not just NVIDIA.

Skeptics note that broadening the capex thesis diffuses the upside narrative. A 2x return on Broadcom is less exciting than a 5x return on NVIDIA, and markets have already priced in much of the Cisco outperformance. The real test is whether margins expand for the ecosystem or whether competition flattens them. Broadcom's gross margin guidance and Cisco's network-fabric pricing power will be the tell in coming quarters.

What to watch next

  • 01Broadcom earnings guidance on fabric-layer pricing power: next week
  • 02Memory-constrained capex trends in next tech earnings: May/June cycle
  • 03Trump-Xi summit impact on approved semiconductor sales to China: ongoing
Mention velocity · last 24 hours
Coverage from these sources
Previously on this story

Related coverage

More about $NVDA

Topic hub
AI Capex: Who's Spending, Who's Earning, and What's at Risk

Tracking AI infrastructure capex — hyperscaler spend, data center buildouts, memory demand and the margin compression risk.