Issue with EVGA 1080ti's and ZCash mining.

[JohnWave]

Hello there!

So I recently received (2) EVGA 1080ti cards and have mixed results.

On my initial attempt I had cards working with the EWBF ZCash Miner version 0.3.4b.

The cards were mining overnight without issue and getting between 650 to 720 Sol/s.

I then decided to add my Radeon 480s on the same rig one at a time until I thought they were stable with a half hour between adding each card.

After adding my third Radeon RX 480 It started crashing. Both the Claymore AMD miner AND the EWBF Miner.

I then decided to remove the Radeon cards from the mix and simply mine with the EVGA 1080ti's.

I was never able to get them functional again to the point where I reinstalled the OS from scratch and STILL was unable to get it working.

I even then looked at other miners with the only success being the excavator_v1.1.0a_Win64 miner.

I let this run overnight and now that too doesn't want to work.

Basically once the miner initializes the screen will blank out and one of the cards fan will spin very high while the other card is idle.

Does anyone have any advise on how I should troubleshoot this?

I am at the point where I just want o go to a Linux build in hopes of getting better, more stable results.

Please advise! Any help is appreciated!

My hardware configuration is as follows:

     BIOSTAR TB350-BTC AM4 AMD Motherboard (6) PCI-E slots
     16GB RAM
     AMD A6-9500 Bristol Ridge Dual-Core 3.5 GHz Socket AM4 CPU
     240GB SSD Hard Drive
     Enermax Platimax PC 1350 Watt PSU (Power Supply)
     (2) PCI-E risers
     (2) NVidia EVGA 1080ti SC2 card

On Windows 10  setup a (16) GB paging file.

[1714782823]

1714782823

[1714782823]

1714782823

[m1n1ngP4d4w4n]

lacking quite a few infos,

drivers version ?
windows version (pro, enterprise, etc...) and patch level (anniversary ? creator update ?) ?
overclock and power levels ?
miner logs of the crash ?

did you tried swapping out the risers to see if you don't have a bad one ?
how do you power the gpu ? (how many cables, do you use Y cables)
did you power the risers properly ?
did you power the motherboard properly with the added molex connectors ?
does one GPU work (one or the other) ? when you put back only one ?

Prob a bad idea to mix up AMD & NVIDIA on the same rig, but i don't think that's your problem anyway.

On a side note, that's way too much SSD & memory, 8GB memory and 128GB SSD would be enough, cpu too powerful too. But that's another matter.

You may also want to increase the virtual memory, make it 16GB start, and 64GB top, it will allow you to run all known algos. (16/32 would do fine too except for a very few niche one like SiB)

[JohnWave]

Answers are below:

drivers version ?

Release Date   Type                   Version           Part Number                  OS                
11-30-2017   Driver           388.43           GeForce 10 Series Family   Windows 10 64bit

windows version (pro, enterprise, etc...) and patch level (anniversary ? creator update ?) ?
Windows 10 Pro

overclock and power levels ?
No overclocking at all - all defaults

miner logs of the crash ?
I set the flag to capture log files but the miner is not creating them...


did you tried swapping out the risers to see if you don't have a bad one ?
I will do that next - I am now troubleshooting with the known good card and it still crashes

how do you power the gpu ? (how many cables, do you use Y cables)
I have (1) 8Pin power cable and (1) 6 Pin power cable in each 1080ti

did you power the risers properly ?
can you please elaborate on this one?
I have the card plugged in to the rise then the 4 pin molex cable is connecting to the rise - the other end of the molex 4 pin is the SATA connector which is plugged in to my PSU thent he USB cable is cabled between the PCI-E slot and the riser. Did I miss anything?

did you power the motherboard properly with the added molex connectors ?
Everything with the computer works great until I start mining.

does one GPU work (one or the other) ? when you put back only one ?
I am now testing with 1 GPU and is still does crash. I also received an

Prob a bad idea to mix up AMD & NVIDIA on the same rig, but i don't think that's your problem anyway.
This is why a rebuilt entirely and now only want to focus on the 1080tis

On a side note, that's way too much SSD & memory, 8GB memory and 128GB SSD would be enough, cpu too powerful too. But that's another matter.
ok


You may also want to increase the virtual memory, make it 16GB start, and 64GB top, it will allow you to run all known algos. (16/32 would do fine too except for a very few niche one like SiB)
I increased the paging file to 32GB

[m1n1ngP4d4w4n]

Answers are below:

drivers version ?

Release Date   Type                   Version           Part Number                  OS                
11-30-2017   Driver           388.43           GeForce 10 Series Family   Windows 10 64bit

Seems ok

windows version (pro, enterprise, etc...) and patch level (anniversary ? creator update ?) ?
Windows 10 Pro

I never tried with W10 Pro, have you patched to latest ? Can you try with enterprise version

overclock and power levels ?
No overclocking at all - all defaults

Ok

miner logs of the crash ?
I set the flag to capture log files but the miner is not creating them...

Ok that's a shame

did you tried swapping out the risers to see if you don't have a bad one ?
I will do that next - I am now troubleshooting with the known good card and it still crashes

Since you have only one card, if you plug it directly on the motherboard without the riser does it work ?

how do you power the gpu ? (how many cables, do you use Y cables)
I have (1) 8Pin power cable and (1) 6 Pin power cable in each 1080ti

Ok seems good, otherwise... i was thinking maybe a faulty PSU, it seems your rig crash as soon as you put it under heavy load, does it crash with only the AMD card ?

did you power the risers properly ?
can you please elaborate on this one?
I have the card plugged in to the rise then the 4 pin molex cable is connecting to the rise - the other end of the molex 4 pin is the SATA connector which is plugged in to my PSU thent he USB cable is cabled between the PCI-E slot and the riser. Did I miss anything? Nope seems good !

did you power the motherboard properly with the added molex connectors ?
Everything with the computer works great until I start mining.

Yes but as soon as you pull the power it's when it crash if not properly powered, so did you connected the additionnal molex connectors on the side of your PCI lanes on the motherboard ?

does one GPU work (one or the other) ? when you put back only one ?
I am now testing with 1 GPU and is still does crash. I also received an

Ok, thinking more & more about the faulty psu.

Prob a bad idea to mix up AMD & NVIDIA on the same rig, but i don't think that's your problem anyway.
This is why a rebuilt entirely and now only want to focus on the 1080tis

Great

On a side note, that's way too much SSD & memory, 8GB memory and 128GB SSD would be enough, cpu too powerful too. But that's another matter.
ok


You may also want to increase the virtual memory, make it 16GB start, and 64GB top, it will allow you to run all known algos. (16/32 would do fine too except for a very few niche one like SiB)
I increased the paging file to 32GB

Cool

[JohnWave]

Here is the latest:

I swapped out out existing risers with new ones. I had an unopened 6-pack of risers that I was saving for my next rig.

I then Only had (1) 1080ti plugged in and its works *perfectly* on its own without issue. In this state it was GPU #0.

I then plugged in the second 1080ti *also* with a new riser and only mined with this one. I set the configuration to only mine with the second GPU which came up as GPU #0 it too worked well on it's own.

I then kicked off in a separate window another instance of the EWBF Miner and specified only to use GPU #1 so GPU #0 and GPU#1 were both running in separate mining instances.

At that point VROOOOOOM!!! - the fans on GPU#1 spun up at 100% and the computer would become unresponsive but the fan on the GPU #1 still keep spinning at 100% until I forced a shutdown of the box.

The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly.

So I am still perplexed but am also leaning towards a PSU issue.

I did look through the Event Viewer and did see this:

     Several of these:  "The system has rebooted without cleanly shutting down first. This error could be caused if the system
     stopped responding, crashed, or lost power unexpectedly."


     "The previous system shutdown at 8:16:18 PM on ‎12/‎7/‎2017 was unexpected."

I then forced both GPUs to mine and here are the Event Viewer entries:

     "The computer has rebooted from a bugcheck.  The bugcheck was: 0x00000124 (0x0000000000000000,
     0xffff8001358458f8, 0x0000000000000000, 0x0000000000000000). A dump was saved in: C:\Windows\Minidump\120717-
     8812-01.dmp. Report Id: 9d4fd139-4e1a-452b-ab24-d8f725a52999."

AND

     "A fatal hardware error has occurred.

     Component: AMD Northbridge
     Error Source: Machine Check Exception
     Error Type: HyperTransport Watchdog Timeout Error
     Processor APIC ID: 0

     The details view of this entry contains further information."

Any ideas?

[wacko]

Any ideas?

Looks like some kind of hardware failure to me. When the fans on a graphics card spin up to 100% instantly it usually means that there's no connection between the card and the motherboard. So in your case the system "loses" the card as soon as you start mining. Not sure why yet, maybe something with the board or the psu. Did you try different pci-e slots?

[m1n1ngP4d4w4n]

Yeah as wacko said, hardware failure, either the PSU or the motherboard, try another PCI-E slot as he said, great idea.

[JohnWave]

Thanks! I will give it a shot!