Nvidia CEO Jensen Huang defended against market reactions to DeepSeek's R1 open-source model, expressing enthusiasm for its potential to boost AI development. He stated that while the announcement led to concerns about a reduction in Nvidia's computing necessity, it actually indicates greater opportunities for efficiency and faster AI adoption. Huang highlighted that despite R1's advancements in pre-training AI models, the post-training phase is crucial and remains a compute-heavy process. Nvidia's stock suffered a 16.9% drop after the R1 release, leading to significant market cap losses, underscoring the tension in the AI industry.
The market got it wrong regarding DeepSeek’s advancements; instead of hindering Nvidia, these developments should accelerate AI adoption, calling the R1 model 'incredibly exciting'.
Despite R1's pre-training breakthroughs, Jensen Huang emphasized post-training will remain resource-intensive and vital for AI, countering claims AI requires less computing.
Collection
[
|
...
]