The major risk would be the rapid tech change. And if you could keep up with the growth speed. Would you be able to keep up with it? And you've done all the research?

gotta run the numbers. it's all about ROI.

You can get the same vram in 2 cards for 1/10 the total cost and get 15 to 20 tok/a. Don't do it.

Just because you aren't finding them on the used market doesent mean their resale value remains high. As soon as the next version comes out, the resale falls like a rock. Electronics is not an investment, it is a capital expenditure.

Bitcoin mining is NOT profitable. It is an expense and those who do it are paying a premium for privacy just like you would be doing with a GPU.

The fact that you can depreciate it as a business expense proves that it does not retain its value. If you expense it, then sell it after its been expensed, anything you sell it for is INCOME and NOT COGS

It seems like you love them and want more of them. The investment part is bs. I know people who like alcohol a little bit too much. They buy and store homemade 70% alcohol and they say it's a store of value. The truth is they will use it all to make various liquors that they will drink themselves an...

Or you could skip the mental gymnastics and buy three ASIC miners instead 😄 At least then the portfolio allocation question becomes much simpler: “how much noise and heat can I tolerate?”

what if the only "help" you got in the comments were generated by gpus?

Rent out the compute for running abliterated models. 🤙

yeah but 20k are still 20k broski