Topic: Mining Reboots - primarily neoscrypt (Read 154 times)

Neoscrypt seems to work the GPU harder than any other algo with the same core/mem settings. For example, a single RX 570 I have mining Neoscrypt is pulling 420W from the wall (the PSU isn't the most efficient) while mining Ethash the power demand is closer to 340W, so 80W less (and mining Cryptonight it is even lower - around 310W).

However, a Seasonic 850W Gold PSU is supposed to be near the top in quality so that doesn't seem a likely culprit here.

I guess hunt through Windows Event Viewer for any errors/warnings.

I replaced the riser on the 1060. but not the 1050.  As I mentioned each card mines individually fine.

The PSU could be a question, but it's a decent PSU. Seasonic 850W Gold, if I recall.

But only neoscrypt? Seems.. like an odd correlation!

You've eliminated overtemp as the proximate cause of rebooting, but that doesn't mean you shouldn't attempt to keep temps at 70C or less.

Next up would be PSU instability from too high and/or too wildly varying of a load.

Also, risers are *always* a suspect when things aren't working right. Try swapping the riser if you have a spare (and you should always have a spare riser if you insist on using them).

I stuck a fan on the rig, along with forcing the GPU fan to 100%

But I was barely mining for 5 minutes and experiences another reboot.

The temperatures didn't even get hot enough to matter yet.

1050: 46C
1060: 61C

[Edit]

Mining Phi for 10 minutes:

1050Ti: 46C
1060:  65C

Something about neoscrypt is making the temperatures higher.

I'm trying to ferret out my own mysterious problem with the display blanking out occasionally so I regret I am not much of an authority on these things, but I can provide perhaps a few things to check:

First, your power supply may not be up to the task even if the sticker on the side says it has sufficient wattage. You might want to buy a Kill-a-Watt to see the real, at-the-wall consumption of your rig.

Second, even 70C is uncomfortably hot for GPUs, and 80C is just asking for core/memory errors; cool those puppies down more.

Third, check Windows Event Viewer for errors/warnings just in case something was reported to it.

Greetings,

 I have a small mining rig.   I'm currently using NemosMiner, and notice that the rig reboots.  I've added some logging, and I've determine that whenever neoscrypt is being mined, I get pretty consistent reboots.

Quick specs:
Win10 (x64)
GPU: 1050Ti
GPU 1060 (3GB model)
Both GPUs are on risers.

I purchased the 1050Ti first, and just ran Neoscrypt straight with it for days.  After adding the 1060, things became shaky, and I also switched over to an algo switcher (Nemosminer)

All that said, I went ahead and ran the mining setup against the 1050Ti standalone for awhile.  As I expected, everything was stable.
Then, I ran the setup against the 1060 standalone. Surprisingly, everything was also stable.

I then combined them again, and that's when things tend to get less stable, and I experience the reboots.

Noteworthy items:
In my troubleshooting attempts I replaced the riser on the 1060 early on, since the 1050Ti was stable before adding it. No help, of course.
I completely removed all overclocking settings.  No help.  
This also appears to happen with Klaust's ccminer, and also palgin's hsrminer.

So, that's where I am.

Physically, the miner is a"milk crate" miner.  Completely open to air flow, and the cards are gapped about 1" apart. There is a desktop fan that blows between them, and it does get warm in there.

I guess my next step would be to separate them further, but that's also not ideal since I would like to eventually put 5 cards in here and they will be near each other.

Would appreciate any further tips to help solve this. Thanks!