Nvidia CEO Jensen Huang says next-generation AI will require 100 times more computing power than older models. This is because new reasoning approaches, like those used in models such as DeepSeek's R1, OpenAI's GPT-4, and xAI's Grok 3, need to think step-by-step to answer questions.
Huang explained that the amount of computation necessary for this reasoning process is 100 times more than what was previously needed. This increased demand for computing power will likely drive sales of Nvidia's AI chips, which are already dominating the market with a share of between 70% to 95%.
Nvidia's latest earnings report shows strong growth, with revenue jumping 78% to $39.33 billion, driven by demand for its AI chips. The company's data center revenue, which includes sales of its market-leading graphics processing units (GPUs) for AI workloads, soared 93% to $35.6 billion.
Huang also highlighted the importance of reducing computing costs to enable the development of more advanced AI models. He noted that Nvidia is working to boost its chip performance every year by two to three times, without increasing costs or energy consumption.
