Jensen's Five-Layer Cake Theory: These Trading Opportunities to Look at!
Next week, NVIDIA GTC 2026 opens its doors. Jensen Huang will take the stage again. Over the past few years, each GTC has served as a major market catalyst.
What will he bring this time? Before the real answers are revealed, let's dive deep into Jensen's most important mental framework — the Five-Layer Cake Theory — and how it can guide us toward investment opportunities in this AI wave.
I. The Five-Layer Cake: From Energy to Applications
Jensen breaks down the AI industrial architecture into five layers, from bottom to top — like a five-layer cake:
Layer 1 · Energy
The foundation of everything. Data center electricity consumption is exploding — nuclear, natural gas, and renewables all benefit. Without stable, affordable, large-scale energy supply, everything else is just talk.
Energy stocks are the most underappreciated beneficiaries of this AI cycle. Nuclear power (CEG, OKLO, TLN) is highly sought after by data centers for its reliable baseload electricity. GE Vernova's power equipment orders are also accelerating.
Layer 2 · Chips
The translators that convert electricity into computation.
TSMC is the 'foundational bedrock' of the chip layer — virtually every advanced AI chip is manufactured there. Broadcom's custom ASIC route is being adopted by Google, Apple, and Meta, creating both a challenge and a complement to NVIDIA.
Layer 3 · Infrastructure
Data centers are AI's factories.
Data center construction is the most capital-intensive segment of this cycle. Oracle's deep partnership with NVIDIA and CoreWeave's rapid expansion both signal that the infrastructure layer opportunity window remains wide open.
Layer 4 · Models
The 'brains' running on the infrastructure. Not just language models — also protein AI, chemical AI, physical simulation models, and autonomous systems. Google, Meta, Microsoft, Amazon, and Alibaba are all heavily invested at this layer.
Layer 5 · Applications
The ultimate point of economic value creation. Drug discovery, autonomous driving, legal copilots, industrial robotics — Salesforce, Shopify, Palantir, and Tesla are all competing here. Every successful application pulls hard on every layer beneath it.
The application layer faces the end market most directly and is where AI commercialization is landing fastest. Tesla's FSD, Palantir's enterprise AI, Salesforce's Einstein AI Agent — all represent real product-market fit.
II. GTC Preview: How Should We Think About NVDA's Stock?
GTC has always been NVIDIA's home turf. Each year, Jensen's Keynote delivers a barrage of new products, new architectures, and new partnerships. Markets often exhibit a clear 'buy the rumor, sell the news' pattern around the event.
Key areas to watch this year:
Blackwell Ultra / Rubin Architecture Progress: The timeline and performance metrics for the next-generation GPU architecture are the most closely watched technical catalyst.
Physical AI & Robotics: Jensen has frequently mentioned embodied intelligence over the past year. Progress on Omniverse and the Isaac platform could create new narratives for the robotics sector.
Software & Services Revenue Mix: CUDA ecosystem, NIM microservices, AI Enterprise — whether NVIDIA's software flywheel is accelerating is critical for valuation re-rating
III. What Do You Think About These Investment Opportunities?
The Five-Layer Cake framework lets us examine the investment opportunity in this AI infrastructure revolution from a full-stack perspective, rather than fixating on the price movement of any single company.
• Which layer are you most heavily positioned in right now? NVDA in the chip layer? Tech giants in the model layer? Nuclear stocks in the energy layer?
• Which layer do you think the market is most underestimating? The energy layer is overlooked by many — yet it's the first principle Jensen emphasizes.
• What moves will you make after GTC? Waiting for the GTC catalyst, or already positioned ahead of it?
Share your thoughts in the comments.
Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.
我认为市场可能低估的层是能源。人工智能数据中心需要大量电力，如果没有可靠的电力，整个人工智能堆栈就无法扩展。Constellation Energy、Vistra Energy和GE Vernova等公司可能会悄然成为人工智能热潮的主要受益者。
至于GTC之前的定位，我更喜欢保持部分定位，而不是追逐炒作。像GTC这样的活动经常会看到“买谣言，卖新闻”，所以我更关注Jensen强调的长期趋势——尤其是AI是否会进一步扩展到机器人和物理AI。🚀
@TigerStars @Tiger_comments @TigerClub
That said, the most underestimated layer is energy and power infrastructure. AI data centres consume enormous electricity, so utilities, grid upgrades, and even nuclear generation could become critical enablers of the AI boom.
The model layer, dominated by Microsoft, Alphabet and Amazon, is already heavily owned, so upside may be more gradual.
For positioning ahead of Nvidia GTC 2026, expectations are already high. A strong Rubin roadmap could extend the rally, but if announcements are incremental, capital may rotate toward AI infrastructure plays such as networking, cooling, and power generation.