Nvidia faces a rotation in institutional flows, with traders turning to Micron and custom AI chips

Nvidia shares have come under pressure as at least some institutional investors appear to be reallocating exposure to other parts of the artificial intelligence chip supply chain, according to an analysis carried by Yahoo Finance. The piece frames the move as a rotation of “smart money” away from Nvidia and toward two adjacent areas: memory and bespoke silicon built for AI workloads.

The Yahoo Finance report points to Nvidia stock weakness alongside activity in Micron shares, suggesting that investors are treating memory as a more immediate lever for AI capacity growth. Micron is widely viewed as a key supplier of the high-bandwidth memory used to feed and accelerate data-intensive AI systems, though the Yahoo analysis does not provide further detail on specific contracts or volumes.

Beyond memory, the report also highlights growing interest in custom AI chips. Custom chips, often referred to as application-specific integrated circuits (ASICs), are designed for specific AI inference or training tasks, and can reduce power use or improve performance versus more general-purpose processors. In the Yahoo framing, that includes the idea that investors are broadening bets beyond Nvidia’s dominant role in discrete GPUs and associated platforms.

The report’s central claim is not that Nvidia is losing all relevance in AI, but that the market’s marginal buyers are shifting. In other words, the same AI buildout that benefits Nvidia may also be creating a larger relative opportunity for companies positioned at different bottlenecks, such as memory bandwidth or the efficiency of purpose-built accelerators.

Nvidia, for its part, continues to market AI data center platforms built around its GPU ecosystem and the software stack that helps customers deploy models at scale. Nvidia has also been actively publishing product and industry updates through its newsroom, where it typically connects new hardware and systems to customer demand indicates. However, the Yahoo Finance article does not attribute its flow observations to any specific Nvidia filing, guidance change, or company communication.

The sector context for this kind of rotation is that AI infrastructure spending is distributed across multiple layers, including compute, memory, networking, and power and cooling. When investors believe one layer is tightening faster than others, they often rotate toward the companies most exposed to that layer. That dynamic is consistent with why analysts and traders can treat the memory supply chain and custom chip development as separate, investable narratives even when the end markets overlap.

What remains unclear from the Yahoo Finance write-up is the level of precision behind the “institutional money flow” framing. The report does not disclose, in the information visible here, the methodology used to define “smart money,” the time window for the trades, or whether the observed flows reflect new longs, profit-taking, hedging, or broader portfolio rebalancing. It also does not tie Micron or custom-chip interest to specific earnings drivers, named customers, or signed supply agreements.

Investors watching next may focus on whether Nvidia’s share performance continues to lag other AI supply-chain names, and whether Micron and custom-chip exposures translate into measurable fundamentals, such as new capacity, gross margin trends, or additional platform adoption. Any new Nvidia commentary on AI infrastructure demand or changes in customer buying patterns would also help clarify whether the current rotation is temporary positioning or a more durable shift in expectations.