Nvidia (NVDA, Financials) shares rose 2.5% to $127.90 as of 12:36 p.m. ET on Thursday, after Morgan Stanley (MS, Financials) reiterated its bullish stance, viewing the recent DeepSeek selloff as a buying opportunity.
Analysts at the company insisted that Nvidia remains in a good position in the AI scene despite mounting worries about competition dangers and long-term investment impediments.
Joseph Moore said that worries about rising export control issues and the growing impact of alternative artificial intelligence systems undermined the opinion surrounding Nvidia. Nonetheless, the company views these advancements as part of the continuous growth of artificial intelligence technology rather than as an existential threat to Nvidia's supremacy. AI hardware has evolved steadily over the last ten years; DeepSeek's arrival is seen as another turn in that cycle rather than a basic change in competitive dynamics.
Emphasizing Nvidia's near-term commercial strength, Morgan Stanley highlighted a better supply situation for Blackwell chips, which are projected to propel major momentum in artificial intelligence training and inference applications. The research highlighted the continuous building of significant clusters, indicating that demand is still intact even though investor worries about a possible halt in the spread of big artificial intelligence training centers have risen. With training workloads being mostly dependent on its GPUs, the company feels that despite recent market fluctuations, Nvidia's supremacy in AI infrastructure is unquestioned.
The research also focused heavily on the long-term development potential in artificial intelligence inference, a field Morgan Stanley expects to propel significant revenue increases for Nvidia over the next few years. The company said that while ASICs have been popular in the industry as a substitute for GPUs for inference workloads, it expects this attitude to change back towards GPUs in the second half of 2025. Critics of ASICs contend that Blackwell, not them, will be the main force for cost cuts in AI infrastructure, thereby negating the belief that ASICs are a better low-cost inference option.
Custom-designed chips are ideal for certain tasks, including artificial intelligence inference, and application-specific integrated circuits provide efficiency benefits over general-purpose GPUs but provide less flexibility for various workloads.
Driven by great demand for its Hopper and Blackwell chips, Morgan Stanley expressed optimism that investor mood will become more bullish despite near-term volatility after Nvidia's next earnings release. With its technology projected to maintain the gold standard for both training and inference workloads, the company regards Nvidia as the unquestionable leader in artificial intelligence acceleration.
