The AI Infrastructure Arms Race Gets Its Own Index
The ACCE AI Infrastructure Index has a NAV of 1009.17, representing a focused bet on companies building the digital backbone of artificial intelligence. Unlike broader technology indices that dilute exposure across consumer apps and legacy hardware, this index zeroes in on firms deriving over 30% of revenue from AI infrastructure — the compute, networking, and data platforms that make AI possible.
Concentration Strategy: Ten Names Control 91% of Assets
The index's structure reveals a high-conviction approach. Ten companies command 91% of total assets, with the top nine holdings each weighted at exactly 9%. This isn't diversification in the traditional sense; it's concentrated exposure to AI infrastructure leaders:
- Cloudflare (NET) - 9%
- Datadog (DDOG) - 9%
- Meta Platforms (META) - 9%
- Amazon (AMZN) - 9%
- Alphabet (GOOGL) - 9%
- Snowflake (SNOW) - 9%
- Nvidia (NVDA) - 9%
- Microsoft (MSFT) - 9%
- Palantir (PLTR) - 9%
- MongoDB (MDB) - 5%
The Hardware-Software-Cloud Trinity
The holdings break into three distinct categories that form the foundation of AI infrastructure.
Compute Layer: Nvidia anchors the semiconductor side, providing the GPUs that train large language models. Microsoft, Amazon, and Google represent the cloud computing platforms where AI workloads actually run. These hyperscalers have spent billions building data centers specifically optimized for AI training and inference.
Data Infrastructure: Snowflake and MongoDB handle the data warehousing and database management that feeds AI systems. Both companies have repositioned themselves as AI-native platforms, with Snowflake's Cortex AI features and MongoDB's vector search capabilities becoming revenue drivers.
Application Layer: Datadog monitors AI applications in production. Cloudflare accelerates and secures AI-powered websites and APIs. Palantir builds AI-powered analytics platforms for government and enterprise customers.
Meta deserves special mention. While known for social media, Meta has become one of the largest investors in AI infrastructure globally. The company's Reality Labs division and massive data center buildouts qualify it under the index's methodology.
Why Equal Weighting Makes Sense Here
The equal-weight structure creates interesting dynamics. Traditional market-cap-weighted indices would heavily overweight Microsoft, Amazon, and Google due to their trillion-dollar valuations. Here, emerging AI infrastructure companies get equal billing with established tech giants.
This approach captures a key insight: the value of AI infrastructure doesn't perfectly correlate with overall market capitalization. Snowflake's specialized data cloud commands higher AI-specific revenue multiples than Amazon's broader AWS platform. Datadog's application monitoring tools become more valuable as AI deployments scale, regardless of the company's smaller total market cap.
The methodology also reduces single-stock risk. Even if one mega-cap technology company stumbles, nine other positions limit downside impact.
Performance Context: NAV Above 1000
The current NAV of 1009.17 suggests the index has generated positive returns since inception (assuming a starting NAV near 1000). This performance comes during a period when AI infrastructure companies faced significant volatility.
2024 and early 2025 saw massive investments in AI infrastructure, followed by questions about monetization timelines. Companies like Snowflake and MongoDB experienced dramatic swings as investors debated whether AI features would drive sustainable revenue growth or merely prevent customer churn.
The index's resilience above 1000 NAV indicates that the equal-weight strategy has weathered this volatility better than approaches that are heavily concentrated in a few mega-cap names.
Missing Pieces and Future Evolution
Notably absent from the top holdings are pure-play AI chip companies beyond Nvidia. Advanced Micro Devices, Intel, and smaller semiconductor firms apparently don't meet the index's revenue threshold criteria. This suggests the methodology prioritizes companies in which AI infrastructure is core to the business rather than a growth segment.
The 30% revenue threshold creates a high bar. Many traditional technology companies offer AI services but derive most revenue from legacy products. The index methodology forces focus on firms that have successfully transitioned to AI-first business models.
Infrastructure Plays for the Long Game
The ACCE AI Infrastructure Index represents a structural bet rather than a momentum play. These ten companies don't just benefit from AI adoption — they enable it. As artificial intelligence moves from experimental to mission-critical, demand for reliable infrastructure grows regardless of which specific AI applications succeed.
The equal-weight approach acknowledges that AI infrastructure leadership remains fluid. Today's dominant platforms may not maintain their positions as the technology evolves. By spreading exposure across the infrastructure stack, the index positions for multiple potential winners rather than betting on a single category or company maintaining permanent advantage.
With enterprise AI adoption still in early stages and consumer AI applications expanding rapidly, infrastructure demand shows little sign of peaking. The companies in this index have built the foundation — now they're positioned to benefit as the AI economy constructs everything else on top.