The Next Wave of AI Growth: Software and Data Over Chips

The AI revolution is no longer just about chips. While semiconductor giants like Nvidia dominate headlines with record revenues, the next phase of growth is being driven by software and data infrastructure. This shift reflects a maturing market where enterprises are moving beyond proof-of-concept AI projects to scale solutions that deliver tangible value. For investors, this means the spotlight is shifting from hardware to the underappreciated software leaders enabling AI's operationalization.

The Software Play: From Data Warehouses to Enterprise AI

Snowflake, once dismissed as an AI "loser," has emerged as a poster child for this transition. Its cloud-based data warehousing and AI solutions, including Snowflake Intelligence, generated a $100 million revenue run rate in the most recent quarter, with 29% overall revenue growth. This reflects a broader trend: enterprises are prioritizing platforms that integrate AI into their data ecosystems rather than standalone tools. Similarly, Salesforce's Agentforce solution, which automates customer service workflows, saw a 330% surge in annual recurring revenue, underscoring the demand for AI that enhances productivity.

Palantir Technologies, often overlooked for its niche focus on data integration, has quietly become a leader in enterprise AI. Its 44% year-on-year revenue growth and updated $2.8 billion full-year guidance highlight its role in solving complex data challenges for governments and corporations. Palantir's platforms, which combine data analytics with AI-driven decision-making, are particularly valuable in high-stakes environments where governance and scalability matter.

MongoDB, a modern database provider, is another standout. Its 21.9% revenue growth in 2025 reflects its appeal to AI applications requiring flexible, real-time data processing. As AI models grow more complex, the ability to manage and analyze vast datasets efficiently becomes a competitive advantage-something MongoDB's architecture delivers.

Market Dynamics: Undervaluation and Long-Term Potential

Despite these successes, software stocks as a whole have faced headwinds in 2025. Fears of AI disrupting traditional software models have led to significant discounts relative to fair value estimates. Morningstar analyst Dan Romanoff argues, however, that many of these companies remain fundamentally sound and offer attractive long-term value. For example, UiPath, which provides AI-powered automation, is trading at a substantial undervaluation according to ValueSense analysis, even as its solutions streamline business processes for enterprises.

The broader AI SaaS market is projected to grow at a 38.28% compound annual growth rate (CAGR), expanding from $71.54 billion in 2023 to $775.44 billion by 2031. This growth is fueled by the fact that 76% of SaaS companies are already using or exploring AI for operations, with 92% planning to increase AI integration in 2025. The key differentiator will be companies that can embed AI into their platforms seamlessly, avoiding the pitfalls of overinvestment in custom models or cloud costs.

Niche Innovators: Data Infrastructure and Enterprise Integration

Beyond the well-known names, niche players are emerging as critical enablers of AI adoption. Applied Digital, for instance, secured a $11 billion AI infrastructure deal with CoreWeave, driving a 132% stock surge in three months. Its ability to build data centers rapidly and secure power for AI research positions it as a hidden gem in the infrastructure space. Dell Technologies, with a forward P/E of 14, is another undervalued player, as AI servers contributed over $16 billion of its $30 billion revenue in a recent quarter.

Shakudo, a newer entrant, is redefining AI infrastructure with its "Operating System for AI on Your VPC." By offering pre-built templates of 200+ open-source tools, it reduces deployment costs and avoids vendor lock-in. This aligns with a broader industry shift toward platform-centric models that unify disparate AI solutions into a single operating layer.

The Road Ahead: Software as the New Bottleneck

While hardware remains foundational, the next bottleneck in AI adoption will be software and data infrastructure. Enterprises are realizing that only about one in three AI implementations delivers measurable ROI, often due to fragmented tools and poor integration. Companies that address these challenges-like Snowflake, Palantir, and Shakudo-are not just surviving; they're thriving.

For investors, the lesson is clear: the next wave of AI growth will be led by software leaders that solve real-world problems, not just those that build faster chips. As the market matures, these underappreciated players will outperform, turning today's undervaluation into tomorrow's returns.