The artificial‑intelligence market is expanding rapidly, and investors are beginning to focus on the infrastructure that supports it rather than on headline‑grabbing consumer applications. While public attention remains on products like ChatGPT, the underlying technologies that enable large‑scale AI models are still in an early growth phase. Three relatively unknown companies appear positioned to deliver returns of roughly 200 % by the end of the decade.
The infrastructure shift that is reshaping the industry
Growth in AI is being driven by the hardware and software layers that allow developers to train and run models at scale. These “picks‑and‑shovels” businesses are comparable to the studios and distributors that supported Hollywood’s golden age, but they operate largely out of the public eye. Unlike well‑known consumer brands, these firms generate recurring revenue from long‑term contracts with enterprises that need reliable, high‑performance compute.
Analysts project the global AI market to reach approximately $1.8 trillion by 2030. Nvidia dominates the graphics‑processing segment, yet a broader ecosystem of equipment manufacturers, data‑center providers, and specialized chip designers is creating the foundation for that growth. The situation mirrors the early days of streaming video, when the most valuable opportunities lay with the companies that stored, delivered, and processed content rather than with the streaming services themselves.
Beyond the hype surrounding conversational models
ChatGPT introduced many investors to generative AI, but the majority of commercial spending is now directed toward enterprise solutions. I have identified three companies that address essential, yet less publicized, challenges in the AI value chain. These firms have solid balance sheets, demonstrated revenue growth, and contracts with large organizations.
The first company provides high‑bandwidth interconnects and custom accelerators that reduce the latency and power consumption of AI workloads. By addressing the “computational bottleneck,” it enables data‑center operators to run more models on existing hardware. Revenue has risen 150 % year‑over‑year, and the firm’s gross margins exceed 45 %.
The second company focuses on efficient memory architectures that allow large models to retain context without requiring petabyte‑scale storage. Its patented approach reduces the energy cost of model inference by up to 30 % compared with conventional designs. Fortune 500 customers have adopted the technology, and the firm’s annual recurring revenue grew 120 % in the most recent fiscal year.
Infrastructure players that the market may be undervaluing
While Nvidia’s stock movements attract frequent coverage, several suppliers that enable Nvidia’s products are also poised for strong performance. These include semiconductor‑equipment manufacturers, data‑center real‑estate trusts, and niche chip designers.
ASML Holdings supplies extreme‑ultraviolet lithography machines—equipment that costs roughly $200 million per unit and is essential for producing the most advanced chips. The company’s order backlog exceeds €30 billion, and its shares trade at a discount to peers despite a technology moat that few competitors can match.
Snowflake operates a data‑cloud platform that consolidates storage, processing, and analytics for enterprise customers. As AI models require ever‑larger training datasets, Snowflake’s usage‑based pricing has generated a compound annual growth rate of 55 % over the past three years, outpacing traditional cloud providers.
The enterprise AI segment with the highest upside
Investing in the infrastructure that powers AI is analogous to owning the studios that produce blockbuster films. The following table summarizes three companies that combine sizable market capitalizations with strong AI‑related revenue growth.
| Company | Market Cap | AI Revenue Growth | Key Advantage |
|---|---|---|---|
| Palantir | $50 B | +40 % YoY | Enterprise AI platform |
| UiPath | $10 B | +30 % YoY | AI‑powered automation |
| ServiceNow | $150 B | +27 % YoY | Workflow automation |
Palantir Technologies has built a suite of platforms—Gotham and Foundry—that help large organizations integrate, analyze, and operationalize data. Commercial revenue grew 54 % year‑over‑year, and the company’s contracts with government agencies and multinational corporations suggest a sizable addressable market that remains underpenetrated.
Since the start of 2023, Palantir’s share price has tripled, reflecting both its expanding customer base and the broader adoption of AI across industries. Analysts estimate that the total addressable market for its services could exceed $200 billion, indicating significant upside potential.
A Chinese player with diversified AI assets
Baidu combines autonomous‑driving technology, cloud services, and a large‑language model known as ERNIE Bot. Its Apollo Go autonomous‑taxi fleet operates in several Chinese cities, providing real‑world data that supports continuous model improvement. Baidu’s AI cloud segment is expanding at a 45 % annual rate, and the company trades at a price‑to‑sales multiple that is lower than many U.S. peers.
The Chinese domestic AI market is projected to reach $70 billion by 2025, according to government forecasts. Baidu’s integrated approach positions it to capture a meaningful share of that growth, offering investors exposure to multiple high‑growth AI sub‑segments through a single ticker.
In summary, the AI revolution is being built on a foundation of specialized hardware, data‑management platforms, and enterprise‑focused software. Companies that supply these components are likely to benefit from sustained demand through 2030. Investors who allocate capital to the three firms highlighted above could see returns comparable to early‑stage technology winners, provided the broader AI market continues on its projected trajectory.