Key points
- Next-generation AI will need 100 times more compute than older models: Nvidia CEO
- Says DeepSeek’s R1, OpenAI’s GPT-4 and xAI’s Grok 3 use a reasoning process
- DeepSeek is fantastic: Jensen Huang
WASHINGTON: Nvidia Chief Executive Officer (CEO) Jensen Huang has said next-generation AI will need 100 times more compute than older models as a result of new reasoning approaches that think “about how best to answer” questions step by step.
“The amount of computation necessary to do that reasoning process is 100 times more than what we used to do,” CNBC cited Huang as saying.
He gave reference to models including DeepSeek’s R1, OpenAI’s GPT-4 and xAI’s Grok 3 as models that use a reasoning process.
AI workloads
Nvidia reported results that topped analysts’ estimates across the board, with revenue jumping 78 per cent from a year earlier to $39.33 billion. Data centre revenue, which includes Nvidia’s market-leading graphics processing units, or GPUs, for artificial intelligence workloads, soared 93 per cent to $35.6 billion, now accounting for over 90 per cent of total revenue.
The firm’s stock still has not recovered after losing 17 per cent of its value on January 27, its worst decline since 2020. That drop came due to concerns sparked by Chinese AI lab DeepSeek that firms could potentially get greater performance in AI on far lower infrastructure costs.
“DeepSeek was fantastic,” Huang said. “It was fantastic because it open sourced a reasoning model that’s absolutely world class.”
Nvidia has been restricted from doing business in China due to export controls that were increased at the end of the Biden administration.
Nvidia counts on billions of dollars of infrastructure spend annually from the largest tech firms in the world for an outsized amount of its revenue. The firm has been the biggest beneficiary of the AI boom, with revenue more than doubling in five straight quarters through mid-2024 before growth decelerated slightly, CNBC reported.
