Jensen Huang predicts AI-generated videos will boost demand for Nvidia chips

Nvidia expects its new AI models – which can create videos and perform human-like voice interactions – to drive demand for the company's graphics processors, CEO Jensen Huang said. “There's a lot of information in life that needs to be backed up by videos and physics, so that's the next big thing,” he told Reuters.

Jensen Huang added: “You have 3D videos and lots of material to learn from. So these systems are going to be pretty big.”

Nvidia's H200 was first used in OpenAI's GPT-4o – a multimodal model that enables realistic voice conversations and can interact via text and images. Nvidia's other customers – Google DeepMind and Meta – have also released platforms for AI image or video generation.

This came after the chipmaker forecast quarterly revenue that far exceeded estimates as it reported more than five-fold revenue growth in its data center unit in the first quarter, sending Nvidia shares up 9 percent.

Derren Nathan, head of equity research at Hargreaves Lansdown, said: “Demand is broad and the large language models will increasingly need to be multimodal, understanding not just video but also text, speech and 2D and 3D images.”

“It (video generation) is certainly one of the strong and already proven use cases for AI and goes beyond just content production,” he said.

In addition, Nvidia's chief financial officer Colette Kress said Tesla has expanded its cluster of processors used in AI training for autonomous driving to around 35,000 H100s.