Billionaire Stanley Druckenmiller Still Isn't Buying Nvidia. But He Just Loaded Up on These 3 Stocks Driving the Next Stages of the AI Boom.

Stanley Druckenmiller bought 195,955 shares of Broadcom, 23,400 shares of Micron Technology, and 411,400 shares of Intel in Q1, while Nvidia remained absent from his portfolio. The moves suggest continued optimism on the AI growth story, especially across chips, CPUs, and memory, even as Nvidia is still seen as the benchmark. The article is mostly positioning commentary rather than new company-specific fundamental news, so near-term market impact should be limited.

Analysis

The important signal is not that these names are “AI beneficiaries” in the abstract, but that capital is rotating from the obvious compute monopoly into the picks-and-shovels layers where pricing power is less headline-driven and more supply-constrained. That usually happens when investors believe the first derivative of AI capex is still intact, but the market is becoming more selective about where incremental returns on capital will accrue. In that regime, the winners tend to be the firms that sit at bottlenecks: custom silicon, memory bandwidth, and legacy CPUs that become necessary as AI workloads broaden beyond training into inference and edge deployment.

Broadcom and Micron are the clearest second-order beneficiaries because their upside is tied to AI-related capacity tightness rather than just unit growth. The key here is that memory and custom ASIC demand can compound faster than the market expects when hyperscalers reallocate budgets from “nice-to-have” experimentation to infrastructure that directly lowers inference cost per token. That creates a setup where earnings revisions can stay positive for several quarters even if AI enthusiasm cools, because supply discipline and customer concentration can sustain margins longer than consensus models assume.

Intel is the contrarian piece: it is less a pure AI winner than a leveraged beneficiary of workload diversification and potential share repair in adjacent server CPU demand. If the market is underestimating how much AI deployment still needs general-purpose compute, Intel can see a multiple rerating from a very depressed base even without becoming a leader in accelerators. The risk is execution slippage; this is a months-to-years story, and any delay in product cadence or foundry progress would quickly re-open the bear case.

The main missing point is valuation dispersion. The crowded trade is still to chase the highest-quality AI platform names, but if capex breadth expands, the better risk/reward may sit in suppliers with less consensus ownership and more room for estimate upgrades. The near-term catalyst set is the next two earnings cycles: if management teams guide to continued supply constraints and AI-related mix improvement, the market should reward these names with both multiple and estimate expansion.

Key Decisions for Investors

• Go long AVGO vs NVDA for the next 3-6 months: thesis is that custom AI silicon and networking attach can rerate on recurring design wins even if the headline AI trade consolidates; risk is that NVDA re-accelerates on new product cycle and regains relative leadership.
• Initiate a tactical long MU position into the next two earnings prints: asymmetric setup if management confirms demand still exceeds supply, with upside from continued gross margin expansion; cut if inventory rebuild signals show up or ASPs compress.
• Take a smaller, higher-risk long INTC call spread 6-12 months out: the trade is for optionality on CPU relevance in AI server buildouts and sentiment repair from a deeply discounted base; risk/reward is attractive only if execution remains intact.
• Pair long TSM / short a basket of high-multiple AI software names over 2-4 quarters: if AI capex broadens, foundry capacity and advanced packaging should capture more durable economics than application-layer valuations, which remain vulnerable to multiple compression.
• Avoid chasing NVDA here unless it retests a materially lower forward multiple; use it as a funding source for relative-value longs rather than a fresh outright add, because upside now depends more on reacceleration than on baseline AI adoption.