Miner- Optiminer/Zcash GPU miner v1.7!

Mining
919

mine is number 0 but i switched it around and no matter what card i used on slot 0 it got dropped after some time
restarting the rig and it will work for a few hours again

920

Since the switch to 0.4.0 Iā€™ve had issues where over night first one card and then the whole rig would lock up (guessing from the hashrates I see in flypool).
Since I switched back to 0.3.4 Iā€™ve had no issues and it has run stable so farā€¦

921

I have had this problem in all versions

922

i could try that switching back to 0.3.4 but now i have it running with claymore with higher hash rate
so unless optiminer puts some effort in it and gets his miner to 15% higher hash then at the moment i see no reason to switch back

923

but only for windows

924

yes took me about 2 hours to get it running on windows 10

925

optiminer hast still a dev hashate of 80k
http://zcash.flypool.org/miners/t1RsS2XVTjEeGST1RdKzPRSu5GTu45WsqcU

so we all need to switch to claymore to make him drop his devfee and get a better own hashrate by the next release.
At all it makes no sense for him to drop his devfee unless his earnings are that high.
He knows also that claymore does not have a linux version and the people are too lazy to switch back to the bad windows OS.

I do not have a problem with a high devfee like this in general, but the overall performance is too low.
I went back to ether, but I will support optiminer again if the hashrate is on claymore level and above.

926

Same, all versions makes me the same.

927

if he has good performance and no issues i donā€™t care what dev fee he takes
but with optiminer dropping gpu and lower hash rate it makes no sense to stay at the moment

929

I am trying to find out where the GPU hangs come from. This is not easy as I do not have a large-scale mining farm and could not reproduce the problem so far. So I need more input from miners.

The hypothesis I have so far: It is either

  • A problem with the risers. Certain host<->device communication patters would then make it more likely to hang.
  • GPU producing wrong results under certain conditions. I am experimenting with different kernels that are more resilient to that.

Did the hangs become obviously worse with 0.4.0? Do they still happen if you reduce GPU clock speed?

930

i would rule out risers as i switched them around and it is always gpu 0 that gets dropped
gpuā€™s are stock

claymore is running without issues so i pretty much blame optiminer

931

I keep my card count at 2 per machine and have had zero problems so far. (Knock on wood)

932

I have many 6 card rigs. Iā€™d be happy to help you test. Stability is extremely important! All my GPUs are Nanos, though, so itā€™s not a mixed-GPU configuration.

933

I believe it is something related to GPU clock speed. I have begun clocking back on GPUā€™s that fall out. Take a look here: http://pariah.ethosdistro.com/

934

Clocking down from the stock clocks?

935

I have a custom mod on my rigs. but for example down clocking core from 1140 to 1070 seems to have made the GPUs stable on the ones that are dropping out.

936

I donā€™t see how down clocking would make sense if a different miner runs without dropping a gpu

937

Stable clocks depend on miner. News at 11.

938

I have tested with different values of clock and memory of the gpu R9 390, the result is the same, restart and freezing of the rig. With Silentarmy does not happen and also with claymores.
regards

939

And I tested on few rigs of r9 390 overclocked to stable maximum (which is lower than for ethereum). Runs fine on Ubuntu 15.04 with flgrx.