May 24, 2024 – According to a recent statement by NVIDIA CEO Jensen Huang, the demand for NVIDIA’s AI chips is expected to surge further with the emergence of new AI models that support video generation. This comes after AI chatbots, represented by ChatGPT, had already sparked a shortage of NVIDIA’s AI chips.
Huang anticipates that the new AI models, capable of creating videos and engaging in human-like voice interactions, will stimulate additional growth in orders for NVIDIA’s AI chips. This prediction aligns with the recent release of OpenAI’s flagship generative AI model, GPT-4o, which boasts text, speech, and video generation capabilities, along with impressive human-like conversational abilities.
In an interview with the media, Huang emphasized the significance of video in everyday life, stating, “There’s a lot of information in life that must be based on video, on physics. So that’s the next big thing. 3D video, and a whole bunch of other stuff, these systems will be very large.”
The increased computational power required to train and run advanced AI systems has driven up demand for NVIDIA’s Grace Hopper chips, such as the H200. This chip was initially utilized in OpenAI’s GPT-4o, a multimodal AI model renowned for its realistic voice conversations and cross-text and image interactions.
Furthermore, NVIDIA’s other clients, including Google DeepMind and Meta, have also unveiled AI image or video generation platforms, contributing to the escalating demand for NVIDIA’s AI chips.
Earlier today, NVIDIA released its first-quarter financial report, revealing a 262% year-on-year increase in revenue and a 628% surge in net profit. Additionally, the guidance for the next quarter exceeded expectations.
According to Derren Nathan, Head of Equity Analysis at Hargreaves Lansdown, “The demand base for NVIDIA chips is broad, and large language models increasingly need to support multiple modes, understanding not just video but also text, speech, 2D, and 3D images.”
Moreover, video AI models employed in the automotive industry are emerging as a significant driver of demand for NVIDIA chips. Colette Kress, NVIDIA’s CFO, mentioned during the earnings call that Tesla has expanded its processor cluster for AI training to approximately 35,000 H100 chips to enable autonomous driving.
Kress also anticipated that the automotive industry would become the largest vertical enterprise in NVIDIA’s data center business this year. Nathan added, “Video generation is absolutely one of the powerful and proven use cases of AI, and it’s not limited to a single content production.”
