Hold onto your wallets, folks! Nvidia’s got the most advanced graphics cards on the market, but it looks like they’re selling for a pretty penny these days. How much, you ask? Try $40,000 on eBay! That’s not a typo, folks. People are shelling out major dough for chips to train and deploy artificial intelligence software.
It was 3D gaming pioneer and former Meta consulting technology chief, John Carmack, who brought this to everyone’s attention on Twitter. On Friday, at least eight of these H100 processors were listed on eBay at prices ranging from $39,995 to just under $46,000. We guess people will do anything for AI technology, even if that means mortgaging their homes.
The H100 is Nvidia’s latest and greatest AI chip, which was announced last year. It’s even fancier than the previous model, the A100, which was about $10,000. Despite its hefty price tag, developers are using the H100 to make large language models (LLMs), which are the basis for AI applications like OpenAI’s ChatGPT. But wait, there’s more! Not only do these systems require tons of data, but they also need some serious computing power to help AI models generate text, images, or predictions.
Training AI models, especially GPT, requires hundreds of high-end Nvidia GPUs working together. What does that mean for us regular folks? Well, it means we should probably stick to our normal jobs and leave the AI stuff to the professionals.
Microsoft, on the other hand, seems to be all-in on this AI craze. They spent hundreds of millions of dollars on tens of thousands of Nvidia A100 chips to help build ChatGPT. We’re not sure how they’re spending their money, but we’re pretty certain that it’s not on housing.
Now, Nvidia’s DGX supercomputer has eight GPUs working together, and it’s all yours for the low, low price of $37,000 a month. Wait, what?! With that kind of money, we could buy a car, a few pairs of designer shoes, and go on vacation. But, if renting access to DGX computers is your thing, go for it!
Nvidia claims that the H100 is the first chip that was optimized specifically for the AI architecture that’s responsible for many of the recent AI breakthroughs. That’s right, they’re coming out with newer, more powerful chips that will be necessary to build even bigger and hungrier AI models. Let’s hope they’ll offer a layaway plan!
Now, go ahead and watch our girl, Google, take aim at Nvidia’s generative A.I!

Serious News: cnbc