Nvidia CEO Jensen Huang has responded to investor concerns after a significant market value drop of $600 billion. Jensen Huang says investors got it wrong over the DeepSeek stock selloff, emphasizing that AI advancements still need powerful chips. The sharp decline was triggered by the release of DeepSeek’s new AI model, R1. Huang argued that the market misinterpreted the model’s impact on AI chip demand.
In an interview for Nvidia partner DDN, Huang stated that investors reacted to R1 as if it marked the end of AI advancements. DeepSeek’s model reportedly used lower-capability chips, raising fears that tech companies might shift away from Nvidia’s high-performance GPUs. The stock drop also impacted Huang’s personal net worth, reducing it by nearly 20%. However, Nvidia’s stock has since recovered most of its losses.
AI Computation Still Demands Power
Despite R1’s efficiency, Huang emphasized that AI models still require substantial computational power. Nvidia maintains that AI development relies on three scaling laws: pretraining, post-training, and test-time scaling. These principles highlight the ongoing need for advanced chips, even as AI models become more efficient.
The launch of DeepSeek’s R1 model caused turbulence in the U.S. tech sector, wiping out $1 trillion from tech stocks. Concerns arose that China might be gaining an edge over the U.S. in AI development. Despite this, major tech leaders, including Google’s Sundar Pichai, Apple’s Tim Cook, and Microsoft’s Satya Nadella, praised DeepSeek’s efforts in their Q4 earnings calls. Jensen Huang says investors got it wrong over the DeepSeek stock selloff, highlighting that DeepSeek’s R1 does not replace high-end GPUs. Nvidia continues to lead the AI chip market, providing high-performance GPUs essential for AI training and deployment. The company has benefited from the ongoing AI boom, with major tech firms competing to secure its chips. Last year, Huang clarified that Nvidia was working to allocate GPUs fairly amid overwhelming demand.
Future of AI and Computational Needs
Huang reiterated that AI’s next frontier lies in improving reasoning capabilities, which remains highly computationally intensive. Nvidia previously highlighted that AI inference requires large numbers of GPUs and high-performance networking. The company remains confident that demand for advanced computing power will persist despite advancements in AI efficiency.
The release of DeepSeek’s R1 model has intensified global competition in AI research. Huang acknowledged the significance of R1, calling its open-source availability an exciting development for the industry.
Analysis of Market Reactions and Future AI Trends
The market’s sharp reaction to DeepSeek’s R1 model reflects the volatility of the AI sector. Investors quickly assumed that the ability to train AI on less powerful chips would reduce demand for Nvidia’s GPUs. However, this assumption overlooks the broader requirements of AI systems, which still rely on high-performance computing for large-scale training and deployment. Nvidia’s response highlights the complexity of AI development, where efficiency gains do not necessarily eliminate the need for powerful hardware.
Despite concerns, Jensen Huang says investors got it wrong over the DeepSeek stock selloff, as AI development requires extensive computational power. AI models are becoming more efficient, but this does not mean that high-end chips are becoming obsolete. Nvidia’s emphasis on test-time scaling and computational intensity shows that AI requires continuous advancements in hardware. The industry is moving beyond simple training models to more complex reasoning-based AI, which demands even more computational resources. While DeepSeek’s R1 model is impressive, it does not change the fundamental need for powerful processing units. Nvidia’s market position remains strong, as AI development continues to require high-performance GPUs for future advancements.
