Topic: Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v15.0 (Windows/Linux) - page 883. (Read 6590565 times)

Yes... i cannot stress enough how important stability is.
ONE single hang or reboot.. and your OC bonus is out the window. Mostly likely you are already in the loss. It might even take a while before you notice unless you have warning systems... and even they take time to activate.

You want them to run for weeks without maintainance.

Ive got the regular black edition and stock speed is 1750. I realize the results will vary but figured 480s could get over 25 easily...at least 27 with any problems occurring.

Thanks. Can it be a dummy plug? I just ordered one for each machine. Do I need one for each card? Or should I just put in in the card I'm having issues with?

thanks again.

Use an hdmi vga adapter on the adapter

I dont think this is a claymore issue, but I have some GPUs which act like they go to sleep and go to 300 mhz. If I hit the "set" button on WattTool it wakes them up and brings they clock up to what it should be.

Any ideas?

What drivers do you use for 980Ti and Win 10 x64 anniversary?

they all give me 3MH/s (even with Ethminer cuda 6.5):

381.65-desktop-win10-64bit-international-whql
368.81-desktop-win10-64bit-international-whql
364.72-desktop-win10-64bit-international-whql
353.49-desktop-win10-64bit-international.hf


there is no 352. drivers for 980Ti

352.63-desktop-win10-64bit-international-beta
352.84-desktop-win10-64bit-international-whql
no not support 980ti

thanks

What brand of card are you using?  What is the stock mem clock speed?

You don't want the highest hash rate where the rig constantly crashes or reboots. You want it to be stable for days and days. You don't make money when it is crashed or rebooting. If the card stops mining then the settings for it are too high. Remember, each card in the same batch can have different results based on the quality of the memory chips (this is just luck). Some can be overclocked more than others. I know a guy that runs a large mining farm. He mods all the BIOS in his cards but never runs them past 1880 MHz for memory speed. He said the stability he achieves is better than pushing them more and having them crashing all day.

Ditto, man.  Rebooting means it's wrong.  Not having to look at it: priceless.

+1 I have followed the exact same logic and performed the exact same actions

still stuck on this, changed to multiple drivers but still cant get over 1950-2000 without my screen going to black. Anyone have this before, would like to get a higher hashrate!

I used to reboot alot, but now that I have figured out that half of the comments on this site are bullshit, (and several are actually quite helpful thankyou), and that 95% of my questions never get answered because they are either difficult or there is no answered, I have quite trying to hit all the unrealistic performance numbers.  So I have tuned all 240 of my cards to run at a stable 27-28 MH/s by lowering the GPU clock speed and downvolting the GPU and the memory to bring my power consumption down to 950W per rig.  Since then I have not rebooted a rig in two weeks except for updating windows over the weekend.  Trying to get super low power or super high hashing rates just makes the rigs unstable and stresses the cards to failure.  I am running all Sapphire nitro+ 470s and 480s, all with mod'd bios's, and all running very well on Claymore's miner now that I stopped trying to set the mining world on fire with higher then sustainable hash rates.  Also, you can use a number of software packages to reset your cards on start up.  Trixx can boot up with start up and reset all the clocks before any other program loads which is what I do to make sure the clocks are reset properly on boot up.  You can still use Watt Tool if you want but you just have to use is in administrator mode.  I still have a few rigs that one card hashes at 7 mh/s and all the other ones are normal (no solution to that?) and I have a few rigs that have a couple cards that hash lower than the identical ones next to it (no solution to that?).  Not really sure the context of my posting except to say keep working at it, as I am sure Claymore is as well.  This is how he makes his living I'm sure, so its in his best interest to give us the most capable miner possible.  Good luck in your efforts.

Keep getting the following


17:49:33:948   1924   Received unknown response: {"id":null,"method":"mining.notify","params":["18f9","0xc05f6fb69d2af63942a2186d040b37347bacbb5ecd2d22c6965ff274ff05868e","0x3bd437c45bff02acb2964c2f09e481e31c61a03eba0cf17d2c30cb05b1ac4d91","0x00000000dbe6fecebdedd5beb573440e5a884d1b2fbf06fcce912adcb8d8422e",true]}

Nope, that one's disabled. Got 2 rigs of 7 GPUs each (2x MSI Z170A Gaming Pro Carbon) and I doubt it'd be wise to enable igfx. The flipside is that the rigs will crash if I connect a monitor to either while they're mining.

Are you using the integrated graphic adapter as your first display?

Thx, I'll check if I can actually install radeon settings with modded drivers... Worth a shot.

Open radeon settings -> gaming -> global settings -> global overdrive, check the reset box and apply ... see if it's still the same. Sometimes the clock and memory remains stuck at a previous value...

Oh and where/why does the temperature monitoring disappear from the console when one uses teamviewer??

Also, trying to set "-cvddc 1000" results in the miner saying "unknown parameter: 1000"
 

...but stability isn't.
Plus, variables do not reset on exit, i.e. You're stuck with whatever values were last set by the miner. Very annoying.

I have noticed people around here tend not to actually solve issues (since they can't be solved by the user) but resort to just reboot the rigs when they fail. I assume the average Claymore-miner user here reboots approx 12 times a day.