do you guys run your different nvidia gpu classes (ie 1070 class, 1080 class etc) on different algo or switching strategies? like 1080tis are better/worse than 1070tis at some algos?
i have 2 470s and 460 that stay at eth on mph, and at the moment my 3 nvidia (1070, 1070ti, and 1080ti) are lumped together as one worker and all mine whatever i choose at mhp. should i split the 1080ti out and run it as its own worker and let it run independent? whattomine seems to suggest that but it varies and from what i gather its not the most reliable thing there is.
jeez life was easier when i was all amd
but i retired my 380/390s and i must admit playing with the nvidia stuff is kinda fun.
i have 2 470s and 460 that stay at eth on mph, and at the moment my 3 nvidia (1070, 1070ti, and 1080ti) are lumped together as one worker and all mine whatever i choose at mhp. should i split the 1080ti out and run it as its own worker and let it run independent? whattomine seems to suggest that but it varies and from what i gather its not the most reliable thing there is.
jeez life was easier when i was all amd
but i retired my 380/390s and i must admit playing with the nvidia stuff is kinda fun.I just got my 1070tis and I intend to run 1080ti and 1070ti differently. 1080tis are dedicated to coins that show linear scaling with cuda cores like vertcoin, skein. Occasionally switch to equihash if there's a pump. 1080ti just don't scale well with equihash.
1070ti probably equihash because it's way more efficient and better price perf in that algo. Though I will switch it to 1080ti's coin if there's a shitcoin pump.
yeah thats the way im leaning too. the 1070/1070ti as one worker, the 1080ti as another. the 1070s for their efficiency, the 1080ti for its grunt. thanks for your thoughts on this.