Nvidia's Huang isn’t shying away from DeepSeek, says AI needs 100x more computing power

Nvidia's Huang isn’t shying away from DeepSeek, says AI needs 100x more computing power

  • 20.03.2025 15:36
  • msn.com
  • Keywords: AI

Nvidia CEO Jensen Huang emphasized that AI requires 100 times more computing power than previously thought, dismissing concerns over China's DeepSeek R1. He highlighted the potential of reasoning AI and Nvidia's new chips, like the Blackwell Ultra, to meet growing demands.

Nvidia ServicesNVDAsentiment_satisfied
open_in_newOriginal article

Estimated market influence

Nvidia

Nvidia

Positivesentiment_satisfied
Analyst rating: Strong buy

Nvidia is leading in AI and benefits from increased computing power demand.

DeepSeek

Negativesentiment_dissatisfied
Analyst rating: N/A

DeepSeek's low-cost model disrupted the market, causing a negative impact on Wall Street.

Context

Analysis of Nvidia's Insights on AI Computing Power and Market Implications

  • Key Assertion by Jensen Huang:

    • Nvidia CEO Jensen Huang emphasized that the world requires 100x more computing power than previously anticipated to handle agentic AI and reasoning AI applications.
    • This shift was described as a "hyperacceleration" in scaling laws for AI, highlighting the growing complexity of AI models.

  • DeepSeek R1 Impact:

    • DeepSeek's R1 model, introduced in late January, caused market disruption by matching OpenAI's capabilities while being trained for only $5 million compared to Silicon Valley's tens of millions.
    • Huang did not express concern over DeepSeek but focused on the need for increased computing power.

  • Nvidia's Strategic Response:

    • Nvidia plans to leverage its new Blackwell Ultra and Vera Rubin superchip to enhance reasoning models, positioning them as critical tools for future AI advancements.
    • The CEO highlighted the role of these chips in advancing physical AI, including humanoid robots and self-driving cars.

  • CUDA and Omniverse Expansion:

    • Nvidia's CUDA software will enable developers to utilize its chips for general processing, reinforcing its dominance in AI hardware.
    • The company is expanding its Omniverse simulation platform, which represents a growing subscription-based revenue stream for metaverse applications.

  • Market Leadership and Long-Term Effects:

    • Nvidia remains the leader in AI technology and is expected to maintain this position due to its focus on cutting-edge hardware and software integration.
    • The emphasis on scalability and innovation suggests long-term growth, potentially leading to further re-rating of its stock multiple as Omniverse matures.

  • Competitive Dynamics:

    • While models like DeepSeek pose competitive pressure, Nvidia's strategic focus on high-performance computing ensures a strong position in the AI race.
    • The company's ability to adapt and innovate will be crucial in addressing the growing demand for advanced AI capabilities.