AI’s Infrastructure Challenges Escalate (Image Credits: Unsplash)
AI data centers could double global power demand by the end of 2026, pushing hyperscalers to invest hundreds of billions in supporting infrastructure.[1][2] Tech giants spent nearly $1 trillion on data center expansions in 2025 alone, creating urgent needs for reliable energy and efficient networking.[2] Two mid-cap companies, Fluence Energy and Credo Technology, have positioned themselves to capitalize on these bottlenecks with innovative solutions for power stability and high-speed data transfer.
AI’s Infrastructure Challenges Escalate
Modern AI training and inference require vast clusters of servers packed into data centers, each consuming electricity on the scale of small cities. These facilities generate immense heat and demand ultra-fast internal networking to handle terabits of data per second. Power grids, often reliant on intermittent renewables, struggle to deliver the clean, steady supply needed to avoid disruptions.
Hyperscalers like Microsoft, Google, and Meta accelerated their buildouts last year, but execution hinges on solving these pain points. Utilities face delays in grid upgrades, while cooling systems strain under higher densities. Companies addressing power reliability and connectivity efficiency stand to gain as capex surges into the trillions over the next few years.[2]
Fluence Energy Delivers Grid-Scale Stability
Fluence Energy specializes in large-scale battery storage systems that smooth out power fluctuations for data centers. Formed as a joint venture between Siemens and AES, the company builds modular batteries capable of gigawatt-hour capacities, ensuring AI servers receive uninterrupted, high-quality electricity even from variable sources like solar or wind.
Revenues remained flat or dipped in 2025 amid broader market softness, but momentum shifted in early 2026. The firm reported strong first-quarter results, reaffirming guidance amid rising data center inquiries.[3] CEO Julian Nebreda stated during the Q1 earnings call, “Accelerating data center growth, utility demand and rising industrial loads continue to drive energy storage demand globally.”[3] Analysts forecast 57% annual sales growth over the next two years, backed by 36 gigawatt-hours of data center projects in development, though not yet contracted.[2]
With a market cap around $1.8 billion as of early 2026, Fluence trades at a premium valuation, reflecting high expectations. Installed capacity reached 46 GWh by late last year, positioning it for outsized gains as AI power needs intensify.[2]
Credo Technology Speeds Up Data Flows
Credo Technology focuses on high-speed connectivity solutions, particularly Active Electrical Cables (AECs) made from copper. These cables transmit data at up to 0.8 terabits per second – matching fiber optics – but use far less power and generate less heat, ideal for cramped data center racks.
Preliminary third-quarter 2026 revenues hit about $406 million, more than tripling from the prior year, with gross margins near 68%.[2] The company showcased next-generation products at the TSMC 2026 Technology Symposium in April, underscoring its role in AI networking.[4] Limited to shorter distances of about 7 meters, AECs suit intra-rack connections where density matters most, helping operators cut cooling costs without sacrificing bandwidth.
Credo’s market cap exceeds $30 billion, driven by hyperscaler adoption, though one client dominated 67% of fiscal 2025 sales.[2] Revenue growth accelerated to 52% in the quarter ending January 2026, signaling sustained demand as data centers scale.[5]
Comparing the Plays: Financial Snapshot
Both stocks offer exposure to AI’s foundational needs, but they differ in scale and risk profiles. Fluence bets on energy storage adoption, while Credo rides networking upgrades.
| Metric | Fluence Energy (FLNC) | Credo Technology (CRDO) |
|---|---|---|
| Market Cap (Early 2026) | $1.8B | $33B |
| Recent Revenue Growth | Flat in 2025; accelerating 2026 | Tripled Q3 2026 prelim. |
| Gross Margin | 11% | 68% |
| Key AI Pipeline | 36 GWh projects developing | Hyperscaler AECs ramping |
Risks include execution delays for Fluence’s unsigned deals and client concentration for Credo. Yet, the AI capex wave – potentially $700 billion annually – provides tailwinds.[6]
As AI models grow more complex, infrastructure will remain the unsung hero of the boom. Investors eyeing the next phase may find Fluence and Credo compelling ways to tap into power-hungry data centers, provided they navigate the volatility ahead.