During a recent online event, Nvidia's (NVDA, Financial) CEO Jensen Huang addressed the market's reaction to DeepSeek's R1 model, emphasizing that the shift in AI training will only increase the demand for Nvidia's computing power. Despite Huang's statements, many investors viewed the situation skeptically, suspecting market manipulation. The U.S. Securities and Exchange Commission has not intervened, leading to speculation about stock price manipulation.
In January, DeepSeek introduced the R1, an open-source inference model, claiming it was developed with less powerful chips and significantly lower investment compared to mainstream Western AI models. This announcement triggered a sell-off of Nvidia's stock, temporarily wiping out $600 billion from its market value and reducing Huang's personal wealth by nearly 20%. However, Nvidia's stock price has since rebounded.
The launch of DeepSeek-R1 raised questions among investors about the necessity of massive investments in AI infrastructure by large tech companies, suggesting that the required computing power for training models might be lower than anticipated. Huang argued that the market's reaction was based on a misunderstanding. Despite R1's development appearing to reduce reliance on computing power, the AI industry still requires substantial computational resources for post-training processing methods.
Huang also highlighted the excitement and innovation that DeepSeek's R1 has brought to the AI field, describing the global enthusiasm as incredible. He suggested that the market reaction could be seen as an opportunity to invest in Nvidia.
