Rig Restarting For No Apparent Reason

Hey guys, some help for a noob please. I’ve only been mining for a day or two and my rig keeps shutting off and restarting. There is no blue screen, no hanging, nothing. It’s as if someone is turning the power off and on again at the wall. A very sudden turn off and then starts back up again. I was just trying to find out why it was doing this when it did it right in front of my eyes and I decided I actually don’t even know where to start other than creating a post here. This happens regardless of whether I use Afterburner or not. In fact, if I don’t set the fan speed to 70% in Afterburner the temps climb way to high (just sub 80°c) If it’s useful, here are my specs:

GPU: 5x MSI GTX 1080 Ti Gaming X 11gb
CPU: Intel Pentium G4000 - 3.50GHz (I couldn’t get the G4560)
PSU: 3000w power supply
Motherboard: ASUS LGA1151 DDR4 B250 Mining Expert
RAM: Ballistix 8gb stick
OS: Windows 10 trial
Miner: EWBF CUDA Zcash Miner

Also I’m not sure if this may be relevant but here’s a screen of my miner where you can see the temp of GPU3 is almost 10 degrees Celcius higher than the others. This only started happening yesterday and again, I’m at a loss as to why!

One last thing…you have disabled sleep mode, correct?

Regarding the temperature disparity, it’s not uncommon. Chalk it up to silicon lottery/differences in heat sink mounting, etc.

Yes I’ve made sure disabling sleep mode was the first thing I did. So just an update as well, the rig has been doing fine so far at 90% Power and 80% fan. The GPUs have been averaging about 65celcius and there have been no crashes. I’ve read a lot about most people needing to not only find the “sweet spot” in terms of Sol/s to Wattage but also having to find the most STABLE configuration. Surely needing to underpower the GPUs in order for it to be stable isn’t normal? Would the fact that the GPUs are brand new and have only been running for a few days be of any importance as well?

Maybe its just me but something does not seem right with that screenshot. The latency, the Sol/w, and the power draw all look way off. I have Evga FTW 3’s pulling better numbers than that. What pool are you using and what server compared to your location? https://imgur.com/AUYaG22

You need to post your config and OC settings

@Edward_Teach I am mining from South Africa and the nearest servers are EU and I’m mining to Nanopool. The normal latency to EU (same with online games) is around 200-250 but the miner fluctuates quite high sometimes.

@CitricAcid My apologies. The OC settings at the time of that screenshot were stock standard except for the fan % being set at a constant 70%. Please tell me where i can find the config you are referring to so i can post it!

Thank you guys for the replies, I’m sure you’ll be able to figure out why this is happening!

config is intended the bat file that you run before mining.
For OC setttings a screenshot of your OC software (afterburner ecc).

As @Edward_Teach mentions, your sol/w efficiency is strangly bad for those cards. You should be able to pull 700 Sol for 200W or less with those cards…

Here is a “good” setting for these cards:

CORE: 1809
MEM: 5258
Driver version: 382.53

(got from this list: ZCash mining GPU Comparison - Wikibooks, open books for an open world)

someone got 735 sol/ for 195W so try to align these values to see if efficiency boosts up.

For your rebooting problem maybe windows trial version that doesnt allow long sessions or long inactivity?

I would suggest you to install a linux distro on a USB and try mining with that and see if reboot is related to HW or SW…

Here is a screen of my OC settings. Regarding the config screenshot, is there anything I need to blank out that shouldn’t be shared with anyone else? I’ve found that these settings are stable, though I am unsure how good they are in terms of output. I don’t know if the temp is the sole cause of the restarts but I’ve noticed that if the temp goes above 70°c the system becomes unstable. With these settings each card is pulling an average of 690 Sol/s.

I figured you might want to see the miner as well:

Regarding the OC settings, my Core Clock only reaches +1000 as well as the Mem. Clock. It seems I just get more and more confused haha.

Edit: I understand you probably meant the overall MHz must be 1809 and 5258, apologies.

After doing my best to get the Core/Mem Clock to be the same as what @johnwisdom recommended I try, here is the result after an hour of it running:

you are still at 100% of power limit, put that at 80% and try adding 400 to the mem clock(+400 in total) and see how those stats are.

:slight_smile:

I ran those stats for a good few hours and the efficiency went up but the avg Sol/s for each card went to around 680-690. I spent most of the afternoon yesterday actually redoing the rig physically to rule out any potential heat issues and even though the Sol/s haven’t gotten much better, it does seem to be fluctuating a lot less and more stable.
Here is my config as well if it helps:

miner --server zec-eu1.nanopool.org --user t1eroyUKX4oHrPcTgE777a6Hzsmf5HZLR7z.Orc_Peon_01/YOUR_EMAIL --pass z --pec --port 6666 --templimit 80

If efficiency went up and you have a more stable situation I would keep this setting, or just tweek around with mem.

For absolute best efficiency try putting power at 75%, core at +150 and memory at +0.

I get 620 sol/s for 145 Watt…

Oh, and temps are crazy low, never over 50 C°, they should last easily many years (> 4)

This sound as a power issue, what brand is your PSU?

Even if it is 3kW it might be low graded and still cause issues ~50% cap!