David Tepper’s Appaloosa trimmed AMD and NVIDIA, while adding to an Amazon AI bet

David Tepper’s Appaloosa Management trimmed exposure to two of the biggest names in AI semiconductors, while boosting a position tied to one of the market’s most widely held AI infrastructure platforms, Amazon. The changes, reported by Yahoo Finance via 247wallst, point to an active portfolio approach rather than a simple buy-and-hold strategy in the AI trade.

Appaloosa reportedly cut its NVIDIA stake by roughly 13% of shares. NVIDIA is a dominant supplier of AI accelerators that power training and inference workloads, and changes in large AI-oriented funds’ holdings are often read as indicates about expectations for demand, pricing power, and the pace of capex among cloud and enterprise customers.

At the same time, the fund also reduced its AMD position by about 32% of shares. AMD is a key challenger in AI compute, supplying both data-center CPUs and accelerators used in AI systems. A larger percentage trim in AMD than in NVIDIA suggests Appaloosa saw more near-term risk or valuation sensitivity in the company’s AI trajectory relative to its leading competitor.

Despite the trims in chips, the report says Appaloosa nearly doubled its Amazon position, adding about 2.14 million shares. Amazon is not a chip manufacturer, but it sits at the center of many AI supply chains through its cloud business. In this kind of portfolio shift, investors are often distinguishing between upstream picks-and-shovels for AI demand, such as cloud capacity, and the semiconductor makers whose results depend on how quickly customers convert spending into orders.

The reported transactions also highlight a broader pattern seen across AI investing. Even when investors are bullish on AI, they may still rebalance between compute components. If cloud operators are believed to keep scaling data-center spending, funds sometimes favor the platforms that receive recurring usage revenue. If, instead, concerns emerge around supply constraints, product competitiveness, or margins among chipmakers, trims can follow even in strong sectors.

Still, the publicly available details in the reported account appear limited to the magnitude of share changes. The report does not lay out what filing period the activity corresponds to, the exact share counts before and after the trades, the average prices at which shares were bought or sold, or whether options or other derivatives were involved. It also does not explain Appaloosa’s specific rationale for AMD versus NVIDIA or why Amazon was increased rather than left unchanged.

For investors tracking the AI trade, the main thing to watch next is whether Appaloosa’s shift aligns with broader market indicates. That includes any changes in the direction of AI infrastructure spending, updates on data-center capex plans from major cloud providers, and comparative performance indicates from AMD and NVIDIA in the AI accelerator market. As always, portfolio updates based on periodic disclosures can lag real-time sentiment, so the market’s interpretation may be less about timing and more about positioning.