The change that Ethereum made in September by switching from “Proof of Work” to “Proof of Stake” significantly affected the earnings of a large group of crypto-miners. As a consequence, all the infrastructure that was used is now looking for a new space in which it can be deemed profitable.
Some of the larger companies that had mines are now ready to replace cryptocurrencies with artificial intelligence. Hive Blockchain and Hut 8 Mining are two companies ready to repurpose their graphics cards for artificial intelligence. Hut 8 in 2022 had earnings of 16.9 million dollars, and part of this amount also goes to calculations for artificial intelligence. Hive plans to increase its revenue 10 times, to $10 million, by 2024.
Both companies made significant investments in hardware, primarily GPUs, in the run-up to the Ethereum switch. They invested tens of millions of dollars in equipment, from which they soon could not turn profit. Some of the big cryptocurrency mines went bankrupt as a result of the aforementioned changes.
“If you can take some of that investment in mining graphics cards and replace them with new cards, then it makes sense,” Hut 8 CEO Jamie Leverton told Bloomberg.
The amount that crypto-miners have invested in equipment is estimated at $15 billion. Repurposing is good news for some infrastructure owners, but unfortunately most graphics cards are not suitable for artificial intelligence. Between 5 and 15 percent of existing cards can be used for AI, the rest will need to be replaced.
This sounds easier than it really is. The price of Nvidia’s Tensor H100 graphics cards is around $10,000. So, replacing unsuitable graphics cards can easily reach super-high sums. Some graphics cards, for example the RTX 4090, can be used for AI, but these account for a small part of the overall hardware.
That AI calculations can help make money is also shown by the fact that every day, for ChatGPT alone, OpenAI spends between $700,000 and $1 million on server costs. This calculation refers only to operational costs and not to the costs incurred for training the artificial intelligence, Business Insider reported.
