The price of the most advanced graphics cards from Nvidia, available for purchase on eBay, has reached more than 40,000.
US dollars, amid high demand for the chips needed to train and deploy artificial intelligence programs.
In a tweet on Twitter, he spotted the 3D gaming pioneer and former director of consulting technology at Meta John Carmack, NVIDIA H100 processor prices.
On Friday, no less than eight H100 processors were listed on eBay with prices ranging from $39,995 to And just under $46,000 USD. Some retailers have in the past offered them for around $36,000 USD.
It is noteworthy that the (H100), which was announced last year, is the latest advanced chip for artificial intelligence from NVIDIA, and it has It came as a successor to the A100 (A100) chip, which sells for about $10,000. They are called processors The backbone of artificial intelligence applications.
Developers use the H100 processors to build what are called large language models, which form the basis of intelligence applications Conversational artificial, such as: ChatGPT chatbot from OpenAl.
These systems are expensive to operate and require powerful computers to transfer terabytes of data for days or weeks at a time.
It also relies on huge computing power so that the AI model can generate texts, images, or Forecasts.
Training AI models, especially large language models such as GPT, requires hundreds of processing units Advanced graphics from NVIDIA, that work together.
Microsoft has spent hundreds of millions of dollars on tens of thousands of i100 chips to help build Chatgpt). NVIDIA controls the vast majority of the AI chip market.
NVIDIA also introduces its DGX supercomputer, which contains eight graphics processing units work together. Earlier this year, the company announced new services that will allow companies to rent access to Computers (DGX) for $ 37,000 per month. At this price, the system will use older (A100) processors nvidia.
NVIDIA says the H100 is the first chip optimized for a specific AI architecture that supports multiple Recent developments in artificial intelligence, which are called Transformers.
Industry experts say more powerful chips will be needed to build larger and more data-intensive models than are available currently.