On Sat, May 4, 2024 at 12:28 PM W2HX via cctalk <cctalk@classiccmp.org> wrote:
> Sure but at $60k per bitcoin it only takes 267 bitcoins to earn a $1M profit > on that $15M per year cost! But I have no idea if a machine like that could > mine 266 coins in a year or 22/month. I'm not really sure if Cheyenne could ever mine a single bitcoin at this point. Bitcoin mining (as I understand it) moved from CPUs (like Cheyenne, even if it has a lot of them) to GPUs many years ago, and it has now even been quite a few years since they abandoned the use of GPUs too. Today it's done on huge clusters of ASIC hash calculators. You need to be able to do something like 3,000 trillion SHA-256 hashes per second to make one Bitcoin per day (at least according to a quick search). If by "Bitcoin" you mean crypto-mining in general, there's probably something it could be used for, but it would be horribly inefficient compared to more modern hardware. The first uses for Cheyenne that seem to pop to mind (Crypto and AI) are actually the two WORST applications for it I think. It might still be reasonably good at the kind of simulations it was acquired to run (atmospheric sims, etc.)
