In This Article:
Nvidia (NVDA, Financials) Chief Executive Jensen Huang said investors misunderstood the significance of DeepSeek's artificial intelligence advancements, contributing to a sell-off that erased $600 billion in market value from the chipmaker before a partial recovery.
In January, DeepSeek, a Chinese artificial intelligence company backed by hedge fund High-Flyer, unveiled R1, an open-source reasoning model. Built on less potent CPUs and with less expenditure than rival Western models, the finding raised questions about the need for massive artificial intelligence infrastructure investments, hence underlining pressure on Nvidia shares.
Emphasizing the still strong demand for artificial intelligence computing, Huang described the response as an overcorrection. While pre-training AI models is crucial, post-training is the key to intelligence, needing major processing capability and sustained demand for Nvidia's processors, he added.
Investor mistrust in artificial intelligence expansion has grown as advancements in AI models seem to be slowing down. For instance, OpenAI has struggled with scalability and added to market uncertainty. According to Huang, the change from training to inference and reasoning points to a change in the direction of artificial intelligence rather than a fall.
Though first wary, Huang said DeepSeek's developments will help the larger artificial intelligence sector. Reiterating Nvidia's view that long-term artificial intelligence expansion is still intact, he addressed questions about model scalability and infrastructure expenditure.
According to Huang's remarks, the ramifications of DeepSeek's advancement for AI investments might be discussed during Nvidia's next earnings call on Feb. 26.
This article first appeared on GuruFocus.