Open newmzy opened 6 years ago
I should have mentioned - I just downloaded the latest pre-compiled binaries, I didn't build from source.
OK, after some reading around some forums I now have the 1060s running nicely, by adding -g 1 -w 64 -I 15 but it seems to only provide me with around about half the hash-rate of the ccminer. Oh well... back to CUDA for me I guess.
Current performance of Maxwells and Pascals is about 70% of the fastest CUDA miner.
Hi - I'm currently using the latest Klaust version of ccminer on 2 mining rigs to mine Feathercoin on a P2Pool node. The rigs are both Windows 8.1, one of them is 4 x gtx1070 and the other is 5 x gtx1060. I was curious to see how your miner would perform compared to the ccminer. I read a kind of 'review' of your miner in which it said that it has 'limited' Nvidia support. If I try to use your example config for a dedicated miner, it always seems to start with gpu0 failed, then proceeds on with each gpu failing in turn. Is it necessary for me to put in the --gpu-platform flag, and if so - what number (arg) should I use? For an idea of what I am achieving now with the ccminer, the 1060s are doing around 4mh/s and the 1070s are at around 4.8mh/s. I would be more then willing to run them on Linux if your miner can do better than that, and is more stable or likely to work in Linux. Thanks for your time.