Building a 2026 Growth Portfolio: 4 High-TAM Plays in the AI Infrastructure Wave

The investment thesis for 2026 is being written in data centers. A massive, secular shift is underway, with global spending on artificial intelligence set to explode. According to market research, global AI spending will jump 44% this year to a whopping $2.5 trillion. Crucially, more than half of that investment is flowing directly into the physical backbone of the AI revolution: infrastructure. AI infrastructure alone accounts for 54% of that outlay. This isn't a fleeting trend; it's a multi-year buildout that creates a towering Total Addressable Market for companies providing the essential hardware and services.

This demand is already fueling a broad-based rally across the AI supply chain. The surge isn't limited to chipmakers. Shares of companies making memory and data storage devices soared in the first weeks of 2026, with some stocks doubling in value as a shortage of critical hardware for training models emerged. The momentum has spread to the firms that design and fabricate the chips themselves, with semi caps like Lam Research and Applied Materials posting gains of 20% or more early in the year. This is a classic "pick-and-shovel" rally, where the beneficiaries are the providers of the fundamental tools, not just the end-product developers.

Yet the market's rotation suggests the rally is still in its early innings. The leadership has shifted dramatically. For the past decade, US large-cap growth stocks-led by the so-called Magnificent Seven-have dominated. But that trend is reversing. In 2025, we saw a partial reversal with international stocks posting their biggest outperformance since 1993, and that momentum has continued into 2026. International outperformance has continued thus far in 2026 and we've also seen outperformance in small caps and value stocks. Meanwhile, the S&P 500 and its growth-focused subsets are down. This broad-based rotation indicates that capital is seeking new sources of growth beyond the established tech giants, potentially opening the door for scalable infrastructure plays to capture a larger share of the AI spending wave.

Stock 1: The Foundational Platform - Nvidia (NVDA)

Nvidia stands as the undisputed bedrock of the AI infrastructure wave. Its graphics processing units have become the de facto standard for training and running AI models, making the company the essential partner for every major data center buildout. This entrenched position is not just about market share; it's about fundamental necessity. As the physical backbone of AI expands, Nvidia's hardware is the single most critical component, creating a massive and durable growth runway.

The company's financial profile underscores its pricing power and operational scale. With a gross margin of 70.05%, Nvidia commands premium pricing while maintaining exceptional efficiency. This margin structure is a hallmark of a dominant platform, allowing the company to reinvest heavily in R&D and maintain a technological lead. The market has already recognized this value, with Nvidia's market cap reaching $4.5 trillion. That valuation, however, is not a ceiling but a reflection of its current dominance. The stock's rolling annual return of 34.5% over the past year shows the market's continued conviction, even as it trades near its 52-week high.

For the growth investor, the key question is not whether Nvidia will participate in the AI wave, but how much of it it will capture. Its position as the foundational platform means it benefits from every new AI application and every new data center project. The prior appreciation is a given; the opportunity now is to assess whether the current trajectory of demand can sustain such a premium valuation. The company's ability to maintain its technological edge and pricing power will be the ultimate test of its scalability in the coming years.

Stock 2: The Specialized Software Layer - Palantir (PLTR)

Palantir is executing a high-stakes pivot from a government-heavy model to a commercial AI software powerhouse. This transition is the core of its growth story for 2026, and the early results are explosive. In the third quarter of 2025, the company's U.S. commercial revenue grew 121% year over year, a staggering acceleration that signals a successful shift in its customer base. This isn't just top-line growth; it's a fundamental expansion of its Total Addressable Market. The customer count is ballooning in tandem, increasing by 45% year over year, indicating the platform is gaining broad enterprise traction.

The engine behind this surge is its Artificial Intelligence Platform (AIP). Palantir's strategy of offering intensive "bootcamp" workshops has dramatically shortened the sales cycle, allowing it to close large deals at a rapid pace. The proof is in the numbers: last quarter, the company closed 204 deals of at least $1 million in value. This commercial ramp is directly benefiting from the broader AI infrastructure spending boom. As companies pour capital into the physical backbone of AI-data centers, chips, and storage-Palantir's software layer is the essential glue that connects enterprise data to generative AI tools, boosting productivity and automation. It's a classic symbiotic relationship where the growth of the foundational layer fuels the demand for the specialized software on top.

For the growth investor, the critical question is scalability and margin expansion. The company is trading at a premium valuation, with a price-to-earnings (P/E) ratio of over 400. That multiple demands that the current hyper-growth trajectory, particularly in commercial software, continues unabated. The company's ability to convert its massive pipeline of large deals into recurring revenue and maintain high gross margins will determine if it can justify its lofty price tag. The setup is clear: Palantir is positioning itself as the indispensable software layer for the AI era, and its commercial numbers suggest it is capturing a significant share of that opportunity.

Stock 3: The Enabling Hardware Builder - Celestica (CLS)

Celestica is the critical, behind-the-scenes builder for the AI infrastructure wave. As an electronics manufacturing services provider, it doesn't make the chips or the servers, but it assembles the complex hardware that connects them. Its role is becoming indispensable, and its stock has rewarded investors for seeing it early. The company's stock soared 177% in the past year, a clear signal that the market is pricing in its outsized growth potential.

The company's explosive growth is directly tied to the AI buildout. Its connectivity and cloud solutions (CCS) segment, which serves the server and storage end markets, saw revenue jump 43% year over year in the third quarter of 2025 to $2.4 billion. This segment now accounts for 76% of its total revenue, making it the primary beneficiary of the surge in data center investment. Celestica's clients are the very architects of the AI hardware stack. The company counts major players in the AI chip industry among its partners, including custom AI chip leader Broadcom, as well as Marvell Technology, AMD, and Intel. It designs and builds the data center networking switches that enable fast connectivity for these custom processors, a function that is in high demand as hyperscalers race to deploy AI workloads.

This client roster provides a powerful moat. Celestica estimates it controlled 55% of custom Ethernet switches last year, a share that grew from 40% the year before. That market dominance, coupled with a new design win for rack-scale data center networking solutions expected to enter mass production in 2026, suggests its growth trajectory is just ramping up. The company's own guidance reflects this optimism, with expectations for a 31% top-line jump and a 39% spike in earnings to $8.20 per share this year.

For the growth investor, Celestica represents a high-beta play on the AI infrastructure trend. It benefits from the same massive spending forecast that drives the sector, with global AI spending predicted to jump 44% this year to a whopping $2.5 trillion. Its business model scales efficiently with that demand, and its current valuation-trading at just 3.2 times sales-suggests the market may still be underestimating the full extent of its participation. The company is positioned to capture a growing share of the custom networking market, turning the broader AI investment surge into exceptional, scalable revenue growth.

Stock 4: The Emerging Capacity Play - IREN (IREN)

For the growth investor, the final piece of the AI infrastructure puzzle is capacity. While companies like Nvidia and Celestica provide the chips and the boxes, a new breed of player is building the physical space and power to run them. IREN is emerging as a pure-play capacity builder, and its recent milestones suggest it is positioning for a major role in the coming wave.

The company's most concrete validation came last year with a landmark deal. IREN secured a 5-year, $9.7 billion deal with Microsoft for 200 megawatts of capacity. This contract is a powerful signal. It demonstrates that hyperscalers are willing to pay a premium for guaranteed, high-quality data center space and the energy to power it, locking in supply years in advance. For a small-cap company, landing such a deal is a transformative event, providing a massive revenue anchor and de-risking a significant portion of its future growth.

Beyond that anchor, IREN's pipeline is what makes it a high-upside play. The company has roughly three gigawatts in its pipeline, a substantial portfolio of projects. The first major site, Sweetwater 1, is a 1.4-gigawatt facility that is ready in April 2026. That timeline is critical. It means IREN is poised to deliver capacity at a time when demand is surging, potentially capturing a large share of the market before many competitors can bring similar scale online. This first-mover advantage in execution is a key part of its growth story.

The investment case here is clear: IREN offers a leveraged bet on the physical buildout of AI. Its stock has already rallied more than 300% over the past year, reflecting early market recognition. Yet, as a company still in its early innings, it trades at a valuation that implies significant future success. The path is not without risk. The company's history as a crypto miner and its current revenue base mean it is still maturing its commercial operations. The execution risk is high-delivering gigawatts of capacity on schedule and securing the remaining pipeline deals are non-trivial challenges. For a growth investor, the trade-off is compelling: the potential for outsized returns from a company capturing a foundational piece of the AI infrastructure wave, balanced against the inherent volatility of a small-cap, early-stage play.

Risks and Counterarguments: Navigating the Growth Wave

The powerful secular trend behind these AI infrastructure plays is undeniable, but it is not without significant headwinds. For the growth investor, the path to dominance is paved with geopolitical friction, intensifying competition, and the high-stakes execution required to convert promise into profit.

Geopolitical risks are a material overhang. The global landscape is shifting, with international stocks posting their biggest outperformance since 1993 in 2025 and continuing that momentum into 2026. This broad rotation away from US large-cap growth suggests a more fragmented and potentially volatile global environment. For companies reliant on complex, cross-border supply chains and manufacturing, this creates uncertainty. The specter of new tariffs or trade restrictions on key markets could disrupt operations and inflate costs, directly pressuring margins for hardware builders and capacity providers alike.

Competition is also intensifying at the very top. While Nvidia remains the undisputed leader, the market is no longer a duopoly. The rise of rivals like Taiwan Semiconductor, which is also in the trillion-dollar market cap club, means the technological and pricing battle is heating up. This competition pressures Nvidia's pricing power and market share, forcing it to continuously innovate and invest at a massive scale just to maintain its lead. For other players in the stack, the presence of such formidable competitors raises the bar for differentiation and customer acquisition.

The highest execution risk, however, lies with the emerging capacity plays like IREN. The company's 5-year, $9.7 billion deal with Microsoft is a transformative anchor, but it is only one piece of a much larger puzzle. IREN's growth hinges on its ability to convert its roughly three gigawatts in pipeline projects into revenue and deliver them on time. The company must navigate permitting, construction, and power procurement at a scale few have attempted. Any delay or cost overrun on its flagship Sweetwater 1 site, scheduled to be ready in April 2026, could undermine investor confidence and its premium valuation. For a small-cap company, the stakes are exceptionally high.

In reality, these risks are the price of admission for participating in a multi-trillion-dollar buildout. The growth story remains intact, but the winners will be those who can navigate these frictions with agility and discipline. The market is rewarding ambition, but it is unforgiving of missteps.

Conclusion: A Multi-Pronged Approach to AI Growth

The AI infrastructure wave is a secular growth story of the highest order, driven by foundational technology that will reshape the economy for years. For the growth investor, the opportunity is to capture this trend not through a single stock, but through a multi-pronged portfolio that spans the entire stack. This approach-investing in foundational hardware, specialized software, enabling hardware, and emerging capacity-allows you to participate in the massive, multi-trillion-dollar buildout while diversifying against the risks inherent in any single segment.

The key metrics to monitor are the pulse points of this expansion. Quarterly revenue growth, particularly for commercial segments like Palantir's AIP or Celestica's connectivity business, will signal the health of demand. Data center sales trends, as tracked by firms like Gartner, provide the macro context; the forecast for global AI spending to jump 44% this year to a whopping $2.5 trillion sets the ceiling for the entire sector. And the competitive landscape is a constant, with shifts like the rise of rivals to Nvidia or new design wins for companies like Celestica indicating who is gaining or losing share in this race.

Catalysts are already in motion. The rollout of new AI chips by Nvidia and its competitors will drive another wave of hardware upgrades and data center deployments. More broadly, the continued expansion of AI adoption across industries-from healthcare to manufacturing-will broaden the base of demand beyond the initial hyperscaler frenzy. For capacity builders like IREN, the catalyst is execution: delivering on their gigawatt pipelines as scheduled.

In the end, the path to outsized returns lies in aligning with this powerful, secular trend. By structuring a portfolio that captures growth at every layer of the AI infrastructure stack, and by watching the right metrics for early signals, investors can position themselves to benefit from the foundational shift that is underway.