Quote from: Homeless_PhD on March 26, 2021, 07:05:27 PM

GPUs are expensive today due to mining profitability - but older GPUs that cant mine ETH OR not well tested (yet) are cheap enough to, potentially, be used to crack #120-125 (at least i think so). I have no money (my budget is ~ 2k$ and thats the money i definitely would not spent now on computers - my "dark\black\doom day" savings) - but some rich enough enthusiast could try to - Tesla K80 it is really powerful ( i assume it should be 30-50% of the Tesla V100 power in BitCrack or Kangaroo). Tesla K80 could be found for 160-300 $. You'll need ~ 400-500 pcs. Each will eat around 300 W that will result in 120-150 kW

) and will require motherboards/cpus/ram/powersource that also cost some money but you could use some old server hardware that is cheap enough (dont know for sure, but expect it to increase the overall cost up to 20-25%, not more). 400 * (300+100)$ == 160 000 $ for hardware AND 3-6 month of 150 kW == 650 000 kWh ~ 65 k$ (at 0.1 $ per kWh). I know - it is still seems not really profitable - but it's fun and the hardware will be able to do some other useful calculations (i believe tesla K80 could be used in mining on some unpopular altcoins - just not well tested/optimized).

150 kW - crazy for home use i'd say (yeah 400-500 GPUs all with server coolers - imaging living on airfield.

(Just reminding - i've tested some Private Keys using my script + BitCrack, you may see in "Examples": https://github.com/HomelessPhD/BTC32)

If you are strong in the construction of graphs try something in geometry to catch (and not lose mind).There, for all sorts of radiuses, the tests of circles, inscribed circles and trapezes, stars etc, the chordam can be calculated something (in theory) https://ibb.co/vq99Sd5 https://en.wikipedia.org/wiki/Bertrand_paradox_(probability)