3x GTX 1060s - 1 with low hashrate issue

New to mining Zcash. I was mining ETH and just recently switched to Zcash.

I am running 3 GTX 1060 6gb but I have one gpu that has a very low hashrate of about 170 Sol/s OCed. The other 2 gpu are getting 300 Sol/s. The only thing I notice is that the one gpu has Samsung memory. Does anyone knows if Samsung is the issue here?

GPU0 GTX 1060 6GB FTW - Micron Memory
GPU1 GTX 1060 6GB FTW - Micron Memory
GPU2 GTX 1060 6GB SSC - Samsung Memory

GPU0 OCed: Pwr 80%, Core +100, Mem +300
GPU1 OCed: Pwr 80%, Core +100, Mem +300
GPU2 OCed: Pwr 90%, Core +100, Mem +500

Fyi, all 3 gpu were running 23 Mh on ETH with no issues.

Thanks for any advice/input.

It’s instructive to post more details when asking for help on a forum. In this case, posting what program and version you’re using to mine ZEC could help. Noting the maker of the cards could be relevant…scant facts help people move on instead of taking time from their activities to help you.

Here’s my solution: get rid of the low performing card and get another like the other two. Pretty obvious solution.

Thanks phakov. Although it is obvious to get rid of the low performance card, I would like to know if anyone is having the same issue with the Samsung memory. I though Samsung memory is better than Micron (based on Eth mining anyway).

Here is what I’m running on.
MOBO- Z270 latest bios
850w Psw
Windows 10 - 1703
Nvidia latest driver
EWBF’s CUDA Zcash miner v0.3.4b
zec-us-west1-nanopool:6666

You should get 280-300 on the 1060. Test it alone on a second PC to see what’s going on to isolate the issue, but getting equal hash power with ETH makes me think it’s a software matter. Maybe updating the drivers could help (???). Try flypool to see if a different pool can help…I use flypool.

I had the same “problem” when I remote desktop to the mining rig the gpu 0 will get lower hasrate because it’s need to uses part of hes power for the remote desktop. Maybe this will also happen when you use the “real” monitor

@brainbudt I have a hdmi dummy connected to gpu0 and the issue was with gpu2. I think I got it working correctly now. After some tweaking and a few reboots later, all 3 gpus are now getting 300+ Sol/s. Thanks all

23 MH for ETH is pretty solid; not a bad idea to stick with that.

300 SoL @ <85 w is the target with 1060’s otherwise.

umm try it without or try to stick in the motherboard and use the igpu being used or something. You could also stop to mine and watch a 4k youtube video and watch witch gpu is working for it.
you could also try to unplug all other gpu’s and try it that way. You need to thinker around I have also sometimes weird problems when I “rebuild” my mining rig.

Hey,

I have the same issue with a 1060 gtx having samsung memory on a Z270 board with Win10.
Also max 170Sol/s for zcash on flypool.
With Ether I get 23-24Mh without any serious oc.

But whatever I do with oc on zcash, I never get past the 170Sol/s

What I already did:
New firmware bios motherboard, set pci things to gen2, lots and lots of oc.
Latest nvidia, older nvidia, no difference.

What did you do to get your card working or maybe any tips (or anyone else) ?

All my cards are running stable now at 310 sol. What I did was re-install Windows 10 uning UEFI. I don’t believe I was using UEFI the first time. The other change I made was PCI Latency in the Bios from 32 to 96.

I can’t remember what driver I’m using at the moment but not the latest driver. Not sure what fixed it but maybe you can try PCI Latency setting first.

unfortunately it did not work with me, I have reinstalled windows10, with uefi enabled.
In the Asus Prime Z270-P X5 motherboard is no such latency settings, so I could not try that.

So I probably go mine some ether instead.

Purchased another 1060 gtx, this time not from MSI, but Asus.

This new card gives me around 305Sol/s without overclock, have to tune down the power to 80% else it would go up to 320Sol/s and I’m afraid the lifespan will not improve by that.

The other MSI 1060 is still slow, but I managed to get it around 175Sol/s.

I think the problem is that this MSI 1060 is in some kind of power-safe mode.
In the Afterburner monitor, the Asus card now uses 80% of its power and the MSI uses Max. 35%, even if I set the maximum to 108% for this MSI.
So probably there is some issue there.

So I go back to ether again, where I get 24Mh/s per card without any crazy oc.