The world's processing power is best served by creating artificial intelligence applications like ChatGPT and google bard than cryptocurrency mining. That's according to Nvidia, which would prefer its GPUs to be used for gaming and AI.
Michael Kagan, Nvidia's CTO, has stated that the company never embraced cryptocurrency and tried to limit its use. RTX-30 graphics chips limit its use for mining.
โAll these cryptographic things, they needed parallel processing, and [Nvidia] it's the best, so people just programmed it to use it for this purpose," he told the guardian (opens in a new tab) . โThey bought a lot of stuff, and it finally collapsed, because it doesn't bring anything useful to society. The AI โโdoes it."
Despite Nvidia's reluctance to engage with cryptocurrency enthusiasts, the company has undoubtedly come out of the Bitcoin and Ethereum bull market well. In all likelihood, he sold a fair amount of GPUs to those who wanted to mine digital currencies.
"I never believed that [crypto] it's something that will do something good for humanity," Kagan added. โYou know, people do crazy things, but they buy your stuff, you sell them stuff. But you don't redirect the company to endorse whatever."
driving the revolution
(Image credit: Nvidia)
While cryptocurrency may be struggling (Ethereum is no longer mineable and the time and energy requirements for Bitcoin are prohibitive), the AI โโrevolution is in full swing. That has been good for Nvidia. According to UBS analyst Timothy Arcuri, OpenAI (the company behind ChatGPT) used 10,000 of the company's GPUs to train the model.
When posing the question of how many Nvidia GPUs were used in its development directly to the chatbot, ChatGPT returned the following response: โThe exact number and type of GPUs used during my training process are not publicly disclosed, but it is known that the development team OpenAI used a large-scale transformer-based language model architecture and trained on a massive text dataset.โ
Currently, tens of thousands of Nvidia A100 and H100 Tensor Core GPUs drive training and inference on AI models like ChatGPT, running through Microsoft's Azure cloud service.
Meanwhile, Nvidia CEO Jensen Huang said it was the engine driving "the iPhone AI moment" at the company's annual conference last week.
More from Tom's guide