AI-Driven Semiconductor Growth

In our latest semiconductor outlook discussion, we broke down the latest shifts in AI-driven semiconductor demand, Nvidia’s evolving role, and what these trends mean for investors. Here are our top three takeaways:

1. AI model development is moving faster than expected: AI model specialization and monetization are accelerating, reshaping semiconductor demand.
2. Hyperscalers are restructuring AI infrastructure: Major players like AWS, Meta, and Microsoft are rapidly optimizing their AI compute strategies, influencing semiconductor investment and CAPEX allocation.
3. The semiconductor opportunity continues to broaden: Nvidia remains dominant, but the evolving AI landscape is creating new demand across GPUs, ASICs, and custom silicon, benefiting a wider range of semiconductor players.

Watch Video
Nvidia Earnings Breakdown and the Next Phase of Semiconductor Growth

1. AI Models Are Scaling and Optimizing at the Same Time

The conversation started with the recent AI breakthroughs from DeepSeek and X AI/Grok. Initially, DeepSeek’s efficiency gains sparked concerns that AI compute scaling might slow. However, Nvidia’s latest results reaffirm that scaling laws are still intact, particularly for US-based operators. The key takeaway? AI still requires significant semiconductor hardware, whether through scaling (GPUs) or optimization (custom silicon and ASICs).

2. Hyperscalers Are Monetizing AI Faster Than Expected

AI monetization has moved beyond expectations, with Meta exceeding forecasts and OpenAI, Microsoft, and AWS carving out distinct AI business models. Rather than competing for the same customers, each is building its own lane, reinforcing demand for different semiconductor solutions. Faster monetization likely translates to higher overall CAPEX spending, benefiting the semiconductor supply chain.

3. Reevaluating Nvidia’s Position in a Broader Semi Landscape

Nvidia remains the dominant AI chip provider, but the discussion highlighted how AI infrastructure is being unbundled. Hyperscalers are ramping up custom silicon efforts (Amazon’s Tranium, Google’s TPU, Meta’s in-house chips), and market share definitions are shifting as AI compute needs fragment. While Nvidia remains a major player, the broader semiconductor ecosystem is expanding, creating more opportunities across different chip architectures. The VanEck Semiconductor ETF (SMH) both offer a diversified ETF approach to capture opportunities across the entire industry.

4. Case Study: ARM and the Acceleration of AI at the Edge

One of the biggest surprises of the quarter was how quickly AI is being integrated into mobile devices. ARM’s licensing revenue has surged as companies like Apple, Qualcomm, and MediaTek adopt its ARMv9 architecture to power AI workloads at the edge. This rapid adoption suggests AI compute is moving beyond just data centers, creating a new growth driver for semiconductors outside traditional GPU markets.

5. The Geopolitical Factor: Intel, TSMC, and US-China Dynamics

The discussion also touched on Intel’s struggles, TSMC’s strategic position, and the long-term implications of US semiconductor policy. With China accelerating its domestic semiconductor efforts and the US relying heavily on Taiwan-based manufacturing, the semiconductor supply chain remains a key area of geopolitical focus.

Final Thoughts from the Past Quarter

The pace of AI innovation continues to exceed expectations, reinforcing demand across various segments of the semiconductor industry. While Nvidia remains a leader, the broader opportunity in AI-driven semiconductors is expanding beyond just GPUs, creating new investment considerations.

As always, our focus remains on mapping these industry shifts as they happen—understanding not just where the market is today but how it is reflexively evolving.

Stay tuned for our next update as we continue to track these developments.