Topic: Claymore's ZCash/BTG AMD GPU Miner v12.6 (Windows/Linux) - page 196. (Read 3839201 times)

i have this problem before... first i found 1 sata power to powering the riser is burning on psu socket... and i think its happen because i use 1 sata wire cable to power 3 rx 480 riser, So i make a change to 1 sata cable to power 2 riser and the problem solved... so check your sata cable power to riser and check every riser if have sign of burn socket

Hi,
I need your help.
I have 2x 280X cards and I was using the v8.0 miner (because that's what the pool I'm using directed me at).
Everything was working fine, until I upgraded to the newest amd driver 2 weeks ago. After that the miner would get stuck in a loop at random times, telling me that there was an issue with the gpu.
I've downloaded v11.1 and tried that, but the problem got worse. The miner will get stuck immediately after startup.
I downgraded my driver to 15.12 hoping that would fix the problem. It works more or less now. I can only use my primary card, in the upper most pci-e slot.
For diagnostics I took that one out and started the miner with the other card in the second slot. I immediately got an error message, telling me that no opencl gpu was found.
Is there a way to get my second card running again? (It will neither work with the v8.0 version that was just running fine before the driver update)

0.53 - 0.61 flypool with 6800

linux 16.04
driver i used the lastest one on their site
same for apk

1125w per rig (6 time rx480)

Can you guys share in which pool how much zec you get for what speed?

I get 0.33 zec/day in dwarfpool with 6200 s/s. Which I think is very low.

With the same speed in coinmine.pl I get 0.52 zec/day.

I have the exact same issue as you describe.

2 x 7990 & 2 x 7950
Enermax 1500w Maxrevo
Celeron cpu
Asrock Z97 Extreme 4

The worst part is it cuts out randomly, sometimes it can be 10mins and others 12hours.

When I have seen the cut outs it happens exactly the same as if the PSU cuts from being over loaded.

I had also observed the power spikes in GPUZ and it normally happens when a share is accepted.

I had guessed that it might cut out, say, in the rare instance that 5 or 6+ shares are found simultaneously which causes a huge spike cutting the PSU which would explain why it cuts after an arbitrary amount of time.

I have no watt meter, however, I assume my cards are running the PSU near to the limit but but I can't check.

I have never really got to the bottom of the problem, however, it seems running Afterburner -20 power limit core at 1050mv leaves my system stable indefinitely.

This is the correct procedure for any issue that is 'strange'.

Start with bare motherboard and min-components to make the computer boot.
Add one card on bare metal (no riser)
Add riser
Add next card on riser etc. etc.

Trying to troubleshoot with 6 cards, 6 risers, and peripherals as possible issues will take 10 times as long.

Why dont just start with one card, observ, put another and etc ? And finally catch the problem

I thought these spikes pretty much disappeared by version 10.x-11? Can someone confirm?

@malandante

looks like you swapped everything out but the actual cards.. and maybe the memory on the mobos? run memtest86+ for a few hours. use the bootable iso on a cd or usb stick, no OS.

set the pcie gen to 2 or 1 in the mobo bios, that seems to keep cards on risers happier.

no bios mods on the videocards? basically everything is stock?

what hashrate has 470 on max intensity and mod bios?and what power consumption?i have now nvidia cards,and i'm interesting about amd qualities in mining because i want define - buy amd or not.

Do you think these spikes can be the reason why my rig freezes? I will describe my problem a little.

I have a 7970, a 280x and two fury nitros. During the last couple of weeks I have been trying to run my system stable with no luck. Freezes can happen at any time after starting to mine. Sometimes minutes, sometimes hours.

Rig specs: Asrock 970 extreme4. Sempron 145. 8GB ram. EVGA 1300 G2. Sapphire 7970 ghz edition, sapphire 280x vapor-x, 3x sapphire fury nitro OC. Tahiti cards undervolted on bios, fiji cards on msi afterburner. No OC on furys. Win 10. 12.6.3. CM 11.1.

First I changed all risers to v6 ones.

Next I upgraded the processor to athlon x2 because I noticed that I was getting less hashes when MSI afterburner was on and I thought the system would be more stable.

After a couple of days more I changed the mobo to asrock h81 pro btc with celeron 1840.

Still no luck.

I thought that the power was insufficient so I added a corsair rx1000 linked with an add2psu device.

No luck, freezes persist.

Changed EVGA to corsair hx1200i. This psu is powering mobo, raisers and tahiti cards.

Nothing changed.

I stopped undervolting. Still same freezes.

One fury died the day before yesterday. No fans, not detected in win 10, not detected on ubuntu, mobo not booting when the card is alone.

I changed risers again to the furys.

I used separate cables for each riser.

I powered the offending fury card with one cable for each connector. Still the same.

I tried catalyst 15.12. Same effect.

Sometimes I see 511 degrees in the fury card on CM log prior to freeze, sometimes it just stops hashing without any error and the watchdog doesn't restart the miner, sometimes all 4 cards go to 0.000 hashing without any error, sometimes after gpu reported incorrect temperature 511°C.

I'm already out of ideas.

Any hint, anyone?

Edit: typo

Hi Guys
Need Your Support
I am having 24 Nos of Power Devil RX470 - 4GB

on Each GPU i am getting 230 sols

and how to increase the sols - i dont know the OC Settings

Help me

Someone posted a link to a Hash-rate/watt-usage table for stock cards a while back. I think it was a link to zcash forum or another Bitcointalk thread. Can someone post it again to refresh my memory?

I will measure it with watt meter and report the results.

I have one of the same cards and it seams no matter what i set the core clock I cant seam to get what most ppl do the crazy thing is if I to set the clocks in AF my rates are lower than if I just start claymore with nothing but temps, it seams that I get the best rates for all my cards this way its just driving me crazy

How many RX480's per Rig? And what is your power consumption?

I get 1770 H/s with x6 RX480's and 950 Watts at the wall

Afterburner/gpu-z most likely show "gpu only" power draw and not from the whole graphics card. So I don't think you're getting 81w at the wall with rx 470, you're getting 100+w.

My bad, I had hd7950 in mind. On rx470 it shows 75w in afterburner so with about 92% efficiency of psu it is about 81w at the wall