Artificial intelligence chipmaker Nvidia has recently increased its dividend and surpassed expectations in both earnings and revenue, boasting adjusted gross margins of an impressive 75%. However, despite these positive indicators, Nvidia’s stock saw a decline in after-hours trading following the release of its quarterly results on Wednesday.
The current phase of AI development, referred to as the “agentic” phase, has shifted attention back to central processing units (CPUs) and more decentralized systems architecture. This transition has led other chipmakers, including Intel, Micron, and AMD, to experience significant stock increases in recent months, while Nvidia’s performance has been comparatively modest.
During the earnings call, analysts sought clarification on Nvidia’s competitive position, especially as the landscape of chip demand evolves. A particular point of interest was the potential impact of Nvidia’s upcoming AI system, Vera Rubin, on its market share in the inference segment of AI technology. CJ Muse, managing director at Cantor Fitzgerald, asked CEO Jensen Huang about Vera Rubin’s influence on Nvidia’s standing in the inference market, which deals with AI’s capability to autonomously answer inquiries and complete various tasks.
Huang emphasized multiple times that Nvidia’s share in the inference market is experiencing rapid growth. He noted, “Our share of inference is growing very quickly,” highlighting the ongoing demand for Nvidia’s technology.
However, Nvidia is not operating in isolation when it comes to AI chip production anymore. The current agentic phase necessitates a shift towards a more distributed processing design, which relies heavily on traditional CPUs instead of the more specialized graphics processing units (GPUs) for which Nvidia has been well-known. This evolving landscape has heightened interest in potential rivals and the competitive dynamics within the industry.
While Nvidia maintains its dominance, the rise of custom silicon offerings from competitors—such as Alphabet’s tensor processing units and Amazon’s Trainium chips—as well as products from AMD and newcomer Cerebras, poses a potential challenge to Nvidia’s market leadership. Despite this, Huang indicated that production capacity for the Vera Rubin chip may be stretched even before its anticipated release later this year, suggesting a strong demand outlook. “My sense is that we’ll be supply constrained throughout the entire life of Vera Rubin,” he remarked.
Furthermore, analysts predict that demand for CPUs and memory is unlikely to diminish anytime soon. Nicolas Gaudois, an analyst with UBS, confirmed this outlook, stating, “Server CPU demand is strongly inflecting up, driving demand… for both conventional server as well as AI servers (CPU head nodes). Agentic AI is a key driver,” pointing to a robust market for Nvidia and its competitors ahead.
