04-13-2019, 09:13 PM
(This post was last modified: 04-13-2019, 09:14 PM by Chick3nman.)
The K80 only has compute capabilities up to 3.7. The number you pointed at is the OpenCL version number and the CUDA runtime version number, not the limited capabilities of the card. The card lacks the LOP3.LUT instruction set introduced in Maxwell and therefor is far less efficient/performant for hash cracking. Choosing the g3 or p3 instances from AWS would be a far better use of your money.