Forget Chips: This Analyst Says 4 “Old School” Themes Could Be the Real AI Industrialization Winners
On CNBC, Kathryn Thompson of Thompson Research Group said AI investors focus too much on near-term chip pricing and should consider “old school” infrastructure themes. She cited $5+ trillion projected AI-related spend over five years and power as a bottleneck. The article highlights Martin Marietta (MLM), CRH (CRH), Ferguson (FERG), Generac (GNRC), and Cisco (CSCO) with recent revenue/EPS and valuation/backlog figures.
How this was made
The 30-second read
Why it matters
The piece uses that thesis to highlight several US-listed industrials with reported quarterly momentum, disclosed deal activity, and (for Cisco and Generac) forward-looking AI infrastructure revenue/backlog signals.
Market read
Traders can use the embedded disclosures (deal size, backlog, and forward revenue guide) to reassess AI infrastructure exposure, but the article is largely narrative and not a single new market-moving event.
What to watch
Investors may focus on margins, capex timing, and execution risk in large projects and acquisitions, which are not quantified here (no cost, integration, or backlog conversion details).
Background
An analyst on CNBC argues investors are over-focused on near-term chip pricing and underweighting a multi-year industrial rebuild behind AI infrastructure.
Ticker impact
Martin Marietta reports Q2 FY2026 results and announces a $13.5B acquisition of Lhoist North America, highlighting capital intensity for AI-linked demand.
Moderate upside bias if deal financing and demand commentary are viewed as credible; otherwise, valuation and YTD drawdown can cap rallies.
The article includes specific deal size and financing plus fresh quarterly metrics, but it is still framed as an analyst theme rather than a standalone market-moving guidance beat.
CRH cites AI infrastructure tailwinds from public investment and reindustrialization, alongside Q1 FY2026 revenue and volume growth.
Limited but constructive reaction potential, mainly as a narrative reinforcement for AI infrastructure demand.
While the quarter numbers are concrete, the piece is primarily thematic and does not disclose a new contract, guidance change, or regulatory event.
Ferguson reports Q2 FY2026 revenue growth and points to durable demand from water infrastructure and large capital projects feeding data centers.
Mild positive bias, with follow-through dependent on broader capex expectations rather than a discrete new order.
The article provides quarterly revenue context but no specific new award or guidance update.
Generac reports Q2 FY2026 revenue growth, a $1.6B data center backlog, and a second hyperscale supply agreement tied to the power bottleneck.
Potential for continued relative strength if investors treat backlog and the hyperscale agreement as incremental evidence of sustained megawatt-generator demand.
Backlog size and the mention of a second hyperscale supply agreement are concrete disclosures, though the article does not quantify incremental revenue impact.
Cisco guides FY2027 AI infrastructure revenue to nearly double to $7.5B, and the article notes shares up over 60% this year.
Support for bullish positioning, with volatility possible if the market already priced the doubling and focuses on execution risk.
A specific forward revenue guide is a primary decision input, but the article does not provide the full assumptions or whether guidance was newly issued on the same day.
Market effects
Reinforces a rotation thesis from chips to power, water, and industrial capex, potentially supporting sentiment for construction materials, distribution, and generator supply chains.
Emphasizes US reindustrialization and public infrastructure spending as the demand driver, which can concentrate flows into US-listed industrials.
If the power and water bottlenecks persist, it can shift global AI infrastructure capex toward regions with faster grid and permitting timelines.
Counterpoint
The article may overstate the immediacy of AI industrialization winners, since power and financing constraints could delay monetization even if demand is real.
Key entities
- analyst/CEOKathryn Thompson
Partner and CEO of Thompson Research Group, arguing AI industrialization is the real multi-year driver versus near-term chip pricing.
- companyMartin Marietta
Reports Q2 FY2026 results and announces a $13.5B acquisition of Lhoist North America.
- companyCRH
Cites AI infrastructure tailwinds from public investment and reindustrialization alongside Q1 FY2026 growth.
- companyFerguson Enterprises
Highlights water infrastructure and large capital projects as durable demand drivers in Q2 FY2026.
- companyGenerac
Reports data center backlog and a second hyperscale supply agreement tied to the power bottleneck.
