NVIDIA’s revenue grew by 262% last quarter, exceeding investors’ high expectations and pushing the company’s shares up in trading. NVIDIA also announced a 10-for-1 stock split effective June 7, and announced a 150% increase in its quarterly dividend.
Revenue for the quarter amounted to $26 billion against previous estimates of $24.7 billion, thanks to record sales of artificial intelligence chips. For the current quarter, revenue is expected to be about $28 billion. NVIDIA shares, which have continued their rapid growth of more than 90% since the beginning of the year, rose 2% per day.
Demand for NVIDIA’s AI-enabled data center GPUs has skyrocketed in the past year as major tech companies rush to develop the computing infrastructure needed to create powerful new AI products. Google, Microsoft, Meta, and Amazon have all indicated that their spending on this will remain high through 2024.
- NVIDIA will now develop new chips every year instead of every two years. This was announced by the company’s CEO Jensen Huang. NVIDIA will also accelerate the development of any other chips it produces to match the once-a-year frequency — that is, we can talk about custom graphics cards
«I can announce that there is another chip after Blackwell. We will work on the basis of one year»,” Huang said.
Previously, NVIDIA created a new architecture approximately every two years — for example, Ampere was presented in 2020, Hopper in 2022, and Blackwell in 2024. Recently, analyst Ming-Chi Kuo reported that the next Rubin architecture will appear in 2025 with the R100 GPU, Huang’s comments indirectly confirm this possibility.
«New processors, new GPUs, new network adapters, new switches… a whole range of chips will appear».
When the analyst asked Huang to explain how Blackwell GPUs will evolve while Hopper GPUs are still selling well, Huang explained that the new generations of NVIDIA chips are electrically and mechanically compatible with previous versions and run the same software. According to him, «customers will easily upgrade from H100 to H200 and to B100» in existing data centers.
Huang also shared a few of his sales pitches during the call, explaining the incredible demand for NVIDIA’s AI GPUs:
We expect demand to exceed supply for a while as we move into H200, as we move into Blackwell. Everyone is looking to connect their infrastructure to the network. And the reason for that is because they’re saving money and making money, and they want to do it as quickly as possible.
Sources: Financial Times, The Verge