Nvidia falls out of favor, AI preferred? RBC analyst calls for buying against the trend

The market has already signaled a cooling trend towards Nvidia. According to Barron's Weekly, although TSMC's impressive financial report briefly boosted its stock price, its core position in the AI industry has weakened. In recent months, funds have clearly flowed into the memory sector - as a core support for AI servers, memory has benefited from the surge in expansion demand, diverting Nvidia's attention. The stock price reflects hesitation more intuitively, entering a consolidation period from October 2025. In early January 2026, it fluctuated in the range of $180-190 and closed at $186.23 on January 16, a slight decrease of 0.44% from the high point at the beginning of the year, forming a contrast with the previous significant upward trend.

The intensification of competition is the key to suppressing market sentiment. Google's self-developed TPU chip accelerates commercialization and directly challenges Nvidia GPU, essentially a game between specialization and generalization routes. Google's seventh generation TPU (Ironwood) has increased computing power by 10 times compared to its predecessor, with energy efficiency 2-3 times that of Nvidia products. It is suitable for the cost reduction needs of large-scale AI clusters and has attracted cooperation from giants such as Meta and Anthropic, aiming to seize 10% of its AI chip revenue. At the same time, AMD's upcoming new chips are expected to carve up the mid to high end market and further intensify market share pressure.

The changes in the industry ecosystem cannot be ignored. The substitution rate of self-developed chips by top cloud manufacturers is rising, and the industry estimates that about 30% of AI inference workloads in the first quarter of 2026 will be borne by self-developed chips, a significant increase from less than 15% a year ago. OpenAI's multi billion dollar collaboration with Cerebras Systems is also seen as a signal to reduce dependence on Nvidia. Combined with the uncertainty of the US chip export policy to China, the sales of H200 chips are subject to strict terms, which compress profit margins and hinder the expansion of the Chinese market. Its market value, which evaporated nearly $100 billion in a single month, reflects the fluctuation of market confidence.

As market doubts heat up, RBC analysts offer a bullish perspective. On January 15th, RBC Capital covered Nvidia for the first time, giving it a "outperform the market" rating and a target price of $240, with a potential increase of over 41%. The Srini Pajjuri team believes that the market exaggerates the threat of competition, and Nvidia's core moat is stable: firstly, the CUDA ecosystem covers more than 95% of AI developers, with high migration costs that are difficult to shake in the short term; Secondly, the strong fundamentals are supported by a backlog of $500 billion orders and a surge in demand for inference; Thirdly, the valuation is attractive, with a discount of about 20% compared to peers and tech giants.

The movement of institutional funds has hidden support. Data shows that 7571 funds or institutions hold Nvidia stocks, an increase of 80 from the previous quarter. Leading institutions such as Pioneer Navigation and JPMorgan Chase have increased their holdings against the trend, with an average allocation weight rising to 3.68%. This indicates that professional investors did not leave due to short-term fluctuations, but instead took the opportunity to increase their holdings, echoing the views of RBC analysts.

The situation of Nvidia is essentially a microcosm of the AI chip industry's transformation from monopolistic growth to differentiated coexistence. JPMorgan predicts that Nvidia's AI chip market share will decrease from 46.5% to around 40% by 2028, still ranking first in the industry; TPU's market share in the field of ultra large scale model training will rise to 15% -20%, forming a binary pattern. For investors, the core is to balance short-term volatility and long-term value. If they recognize the high growth of AI and Nvidia's advantage, the current consolidation period is a good opportunity for layout. Otherwise, they need to be alert to the risk of valuation correction.