Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v15.0 (Windows/Linux)

[Call_Me_Bambi]

I mine 15 days in nicehash and 15 days with claymore and other miners sometimes. I like to mine btc because that's the btc I use to pay bills and other stuff I manage to pay with the amount I'm able to mine the 15 days. Then I like to play or trade the coins that I mine for the rest of the month. I know that you should only care to mine the most profitable coin but I like to take advantage of early mining because the rewards are shared among less miners and where there's no way of getting them other than mining, I like to take the risk of giving the new coin a chance.

But each person will have a different strategy for their mining.

Personally I would like to hold what I mine. I'm getting into mining this month with a rig and I don't know if I should do nicehash and get BTC or Claymore and get ETH. BTC is tons more stable but it doesn't have the potential value increase of ether I think.

Forget both Ethereum and Bitcoin and mine Sumokoin https://bitcointalk.org/index.php?topic=1905086.0 , https://pool.sumokoin.com/ .
It uses the Cryptonote algorithm which takes 25% less electricity costs (some report 30% less) than the Ethash algo, and cards run cooler.
If you really want to hold BTC or ETH then just mine Sumokoin (Cryptonote algo) and convert to BTC or ETH.
Claymores Cryptonote miner: https://bitcointalk.org/index.php?topic=638915.0
Sgminer for Cryptonote: https://github.com/genesismining/sgminer-gm/releases <--- I prefer this one for Cryptonote.

[bambinn]

Every time i find a Sia coin share (i am using dual-mining ETH+SIA on nanopool) i get a recieved error:

Code:

SC: 11/11/17-17:53:13 - SHARE FOUND - (GPU 0)
SC: Received error: {"jsonrpc":"2.0","id":4,"error":{"code":-32700,"message":"target-miss"}}

And "Current Calculated Hashrate" (for SIA) on Nanopool is always 0, "Average Hashrate for last 6 hours" is however NOT 0, and i do seem to mine a few SIA coins now and again.

Note! Eth mining is working fine!

This is the bat-file i use to load up Claymore Miner:

Code:

setx GPU_FORCE_64BIT_PTR 0
setx GPU_MAX_HEAP_SIZE 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_SINGLE_ALLOC_PERCENT 100
EthDcrMiner64.exe -epool eth-eu2.nanopool.org:9999 -ewal MYETHWALLET.MYWORKER/MY@EMAIL -epsw x -dcoin sia -dpool stratum+tcp://sia-eu2.nanopool.org:7777 -dwal MYSIAWALLET/MYWORKER/MY@EMAIL -dpsw x -ftime 10

Any help would be much appreciated!
Thx in advance!!

[bardacuda]

halp! I am trying to mine on linux but I get the error "No AMD OPENCL or NVIDIA CUDA GPUs found, exit"

I have an R9 270 and an RX 570 in the system, but I am only trying to mine on the 570. I have a few different opencl libraries installed and the AMD-APP-SDK v3.0.

Below is the output from lspci -v and dpkg -l as well as my mining script. What am I doing wrong?

#!/bin/sh
export GPU_FORCE_64BIT_PTR=0
export GPU_MAX_HEAP_SIZE=100
export GPU_USE_SYNC_OBJECTS=1
export GPU_MAX_ALLOC_PERCENT=100
export GPU_SINGLE_ALLOC_PERCENT=100
./ethdcrminer64 -epool us1-etc.ethermine.org:4444 -ewal <MyWalletAddress>.<WorkerName> -epsw x -platform 1 -esm 1 -etha 0 -asm 1 -ethi -49152 -dcri 24 -eres 0 -mode 1 -mport 0 -allpools 1 -allcoins etc -di 1 -cclock 1220 -mclock 2050 -cvddc 900

https://i.imgur.com/tpWBjYp.png

linux peeps halp meh plezz!

I've gotten slightly further along by installing a bunch of the .deb packages from the amdgpu-pro 17.40 tarball and adding this line to my script:

LD_LIBRARY_PATH=/opt/amdgpu-pro/bin/clinfo

But now I get "Cannot build OpenCL program for GPU 0/Segmentation fault" and also it only sees 3GB of VRAM. This is the same error I was getting on Ubuntu 16.04 (I'm on Debian Buster now).

https://i.imgur.com/t1e5W08.png

EDIT: nvm I got it hashing. Had to install OpenCL from the ROCm github. Also had to run Claymore without the script and just add the parameters to config.txt for some reason or I would still get the OpenCL error.

Could use some help getting hash rate up though. I'm still getting pre-dag-fix hashrates. I already added "amdgpu.vm_fragment_size=9" to the /etc/default/grub file so not sure what else to do there.

Also getting "ADL Library not found." I downloaded and compiled AMD_ADL_SDKV10 and just got an "adlutil" binary, which also gives the same error when I try to run that.

[EthereumBa]

Hi guys. I'm getting this error since yesterday. Trying to fix it but no success for now. Any hints what could be a problem?

Photo: https://www.photobox.co.uk/my/photo/full?photo_id=21867640249

GPU1, GpuMiner cu_k01 failed 73, an illegal instruction was encountered
GPU4, GpuMiner kx failed 1
GPU1, GpuMiner kx failed 1
GPU 4 failed
GPU 1 failed
Setting DAG epoch #150 for GPU0
GPU0 failed
GPU0, CUDA error 73, cannot write buffer for DAG
GPU 0 failed
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
WATCHDOG: GPU error, you need to restart miner

I have an 6x GTX1070 STRIX miner. Core clock +99, Memory clock +600. Tried to decrease it, but no effect for now. Running 10.0 Claymore Dual eth+sia. Power is on 90% for now. Tried to change it but also no effect.

Have you previously make this worked and suddenly have an error?

seems like a driver error or miner config problem to me

Yes the rig was working about 3 days without an error at +150 core clock and +650 memory clock. But I haven't touched the config file. What should I edit inside?

I got a stable build with memory clock decreased to +540, power to 100% and the last Claymore 10.1. I still got the "incorrect share" errors but its working stable for 34 hours now.

Maybe the problem is the PSU because I got an Corsair HX1200i Platinum for my build. I'am not home at the moment, so in a month I'll buy a second PSU and then try to raise the clocks again. Still I got an average 182MH/s which is not bad. But I think I can go to 200MH/s.

[ZiDanRO]

For the ones who are complaining about force restarts and freeze i will tell you my experience, i had the same problems and maybe this will help you to:
I have 5 sapphire rx580 8GB and 1 msi rx580 gaming edition 8GB, windows 10 64bit ver 1703
1. Until robin hood drivers i didn't have any problems with claymore or drivers or freeze. After installing this drivers i could not overclock my cards from claymore so i needed to tweak them with Afterburner and BIOS edits, but again i could stil function properly
2. When the official drivers from amd (still beta, but from official site) where launched i installed them and every 2-3 days I needed to reboot my sister. No matter what i have tried: OC tweets (even with the same settings which previously worked), I`ve seen the spikes in energy drawn by the sistem like Claymore said is causing the freeze.
3. Tha final step to create a stable sistem:
 - uninstall drivers with AMD wizard which installs also the 17.11.1 drivers. I didn't need to uninstall them from safe mode
 - patch the drivers
 - uninstall Afterburner
 - installed Claymore 10.1 with same Config file which worked with Robin Hood drivers
 - set every card in Compute mode, reboot the sistem and everything was set.

I already have a week without freeze or reboots.
Claymore could set everything except the fan speed of my MSI card: cclock, mclock, cvddc, mvddc
With 17.11.1 drivers i observe also a 50-60w economy on my sistem consumed

I hope this will help also others!

[EthereumBa]

Hi guys. I'm getting this error since yesterday. Trying to fix it but no success for now. Any hints what could be a problem?

Photo: https://www.photobox.co.uk/my/photo/full?photo_id=21867640249

GPU1, GpuMiner cu_k01 failed 73, an illegal instruction was encountered
GPU4, GpuMiner kx failed 1
GPU1, GpuMiner kx failed 1
GPU 4 failed
GPU 1 failed
Setting DAG epoch #150 for GPU0
GPU0 failed
GPU0, CUDA error 73, cannot write buffer for DAG
GPU 0 failed
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
WATCHDOG: GPU error, you need to restart miner

I have an 6x GTX1070 STRIX miner. Core clock +99, Memory clock +600. Tried to decrease it, but no effect for now. Running 10.0 Claymore Dual eth+sia. Power is on 90% for now. Tried to change it but also no effect.

Have you previously make this worked and suddenly have an error?

seems like a driver error or miner config problem to me

Yes the rig was working about 3 days without an error at +150 core clock and +650 memory clock. But I haven't touched the config file. What should I edit inside?

I got a stable build with memory clock decreased to +540, power to 100% and the last Claymore 10.1. I still got the "incorrect share" errors but its working stable for 34 hours now.

Maybe the problem is the PSU because I got an Corsair HX1200i Platinum for my build. I'am not home at the moment, so in a month I'll buy a second PSU and then try to raise the clocks again. Still I got an average 182MH/s which is not bad. But I think I can go to 200MH/s.

I restarted it today at 36 hours, and now its crashing every 3 minutes. Its driving me nuts!

EDIT: I lowered the memory clocks to +530, now it's working again.

[leonix007]

Hi guys. I'm getting this error since yesterday. Trying to fix it but no success for now. Any hints what could be a problem?

Photo: https://www.photobox.co.uk/my/photo/full?photo_id=21867640249

GPU1, GpuMiner cu_k01 failed 73, an illegal instruction was encountered
GPU4, GpuMiner kx failed 1
GPU1, GpuMiner kx failed 1
GPU 4 failed
GPU 1 failed
Setting DAG epoch #150 for GPU0
GPU0 failed
GPU0, CUDA error 73, cannot write buffer for DAG
GPU 0 failed
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
WATCHDOG: GPU error, you need to restart miner

I have an 6x GTX1070 STRIX miner. Core clock +99, Memory clock +600. Tried to decrease it, but no effect for now. Running 10.0 Claymore Dual eth+sia. Power is on 90% for now. Tried to change it but also no effect.

Have you previously make this worked and suddenly have an error?

seems like a driver error or miner config problem to me

Yes the rig was working about 3 days without an error at +150 core clock and +650 memory clock. But I haven't touched the config file. What should I edit inside?

I got a stable build with memory clock decreased to +540, power to 100% and the last Claymore 10.1. I still got the "incorrect share" errors but its working stable for 34 hours now.

Maybe the problem is the PSU because I got an Corsair HX1200i Platinum for my build. I'am not home at the moment, so in a month I'll buy a second PSU and then try to raise the clocks again. Still I got an average 182MH/s which is not bad. But I think I can go to 200MH/s.

I restarted it today at 36 hours, and now its crashing every 3 minutes. Its driving me nuts!

Do basic isolation, exclude all other GPU's and install only one in PCI-ex16 slot without riser, this is to mitigate PSU issue and riser issue.

if nothing happens, then second step is to clean uninstall drivers and install, we are talking about only 1 GPU here, make work with only 1 gpu when it stabilized then install the other GPU's

[crypper]

Hi guys. I'm getting this error since yesterday. Trying to fix it but no success for now. Any hints what could be a problem?

Photo: https://www.photobox.co.uk/my/photo/full?photo_id=21867640249

GPU1, GpuMiner cu_k01 failed 73, an illegal instruction was encountered
GPU4, GpuMiner kx failed 1
GPU1, GpuMiner kx failed 1
GPU 4 failed
GPU 1 failed
Setting DAG epoch #150 for GPU0
GPU0 failed
GPU0, CUDA error 73, cannot write buffer for DAG
GPU 0 failed
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
NVML: cannot get current temperature, error 999(an internal driver error occurred)
WATCHDOG: GPU error, you need to restart miner

I have an 6x GTX1070 STRIX miner. Core clock +99, Memory clock +600. Tried to decrease it, but no effect for now. Running 10.0 Claymore Dual eth+sia. Power is on 90% for now. Tried to change it but also no effect.

Have you previously make this worked and suddenly have an error?

seems like a driver error or miner config problem to me

Yes the rig was working about 3 days without an error at +150 core clock and +650 memory clock. But I haven't touched the config file. What should I edit inside?

I got a stable build with memory clock decreased to +540, power to 100% and the last Claymore 10.1. I still got the "incorrect share" errors but its working stable for 34 hours now.

Maybe the problem is the PSU because I got an Corsair HX1200i Platinum for my build. I'am not home at the moment, so in a month I'll buy a second PSU and then try to raise the clocks again. Still I got an average 182MH/s which is not bad. But I think I can go to 200MH/s.

I restarted it today at 36 hours, and now its crashing every 3 minutes. Its driving me nuts!

EDIT: I lowered the memory clocks to +530, now it's working again.

Have you measured power from the wall? With your clocks and 100% TDP I bet you have psu problem (just extrapolating from my clocks and tdp as I run at 70%).

[EthereumBa]

Now its crashing again with +530 memory clocks.

I'm not home at the moment, so all the physical stuff will wait for some time now. How can I measure the power from the wall, with a multimeter?

I will try a clean driver install. First uninstall them with DDU, then install fresh ones again.

[MikeGR7]

Now its crashing again with +530 memory clocks.

I'm not home at the moment, so all the physical stuff will wait for some time now. How can I measure the power from the wall, with a multimeter?

I will try a clean driver install. First uninstall them with DDU, then install fresh ones again.

Brother, brother, brother...

Let me help you a bit.  

First of all i miss critical information from your setup.

You have nVidia so the Nr.1 thing i need is to tell me what type of memory chips is in your cards. Check GPUZ and tell me.

Nr.2 you must do is this: Stop installing and uninstalling drivers cause it's useless.
Just go in nVidia control panel and disable any power saving from 3D settings. Choose max performance mode.

Nr.3 Set Windows Power plan in high performance, no sleep, no turning off display. Set 20000MB Virtual Memory, disable scheduled disk optimization, win updates and antivirus scheduled scans (Leave antivirus ON).

Last thing go to Afterburner and reset all settings, only thing you change is fan control to max fan speed above 67C.

After all that run miner again and report me if it's stable and we can work from there  

[EthereumBa]

Now its crashing again with +530 memory clocks.

I'm not home at the moment, so all the physical stuff will wait for some time now. How can I measure the power from the wall, with a multimeter?

I will try a clean driver install. First uninstall them with DDU, then install fresh ones again.

Brother, brother, brother...

Let me help you a bit.  

First of all i miss critical information from your setup.

You have nVidia so the Nr.1 thing i need is to tell me what type of memory chips is in your cards. Check GPUZ and tell me.

Nr.2 you must do is this: Stop installing and uninstalling drivers cause it's useless.
Just go in nVidia control panel and disable any power saving from 3D settings. Choose max performance mode.

Nr.3 Set Windows Power plan in high performance, no sleep, no turning off display. Set 20000MB Virtual Memory, disable scheduled disk optimization, win updates and antivirus scheduled scans (Leave antivirus ON).

Last thing go to Afterburner and reset all settings, only thing you change is fan control to max fan speed above 67C.

After all that run miner again and report me if it's stable and we can work from there  

In the meantime I played around again and set the memory clocks to +530, and power to 95%. Now its running for 2 hours without crashes.

1. The cards have Micron memory.
2. Done. Power management mode - Prefer maximum performance.
3. All of that I have done before I started mining.
4. I can turn on the fan speed, but the cards never got over 62C.

[MikeGR7]

Now its crashing again with +530 memory clocks.

I'm not home at the moment, so all the physical stuff will wait for some time now. How can I measure the power from the wall, with a multimeter?

I will try a clean driver install. First uninstall them with DDU, then install fresh ones again.

Brother, brother, brother...

Let me help you a bit.  

First of all i miss critical information from your setup.

You have nVidia so the Nr.1 thing i need is to tell me what type of memory chips is in your cards. Check GPUZ and tell me.

Nr.2 you must do is this: Stop installing and uninstalling drivers cause it's useless.
Just go in nVidia control panel and disable any power saving from 3D settings. Choose max performance mode.

Nr.3 Set Windows Power plan in high performance, no sleep, no turning off display. Set 20000MB Virtual Memory, disable scheduled disk optimization, win updates and antivirus scheduled scans (Leave antivirus ON).

Last thing go to Afterburner and reset all settings, only thing you change is fan control to max fan speed above 67C.

After all that run miner again and report me if it's stable and we can work from there  

In the meantime I played around again and set the memory clocks to +530, and power to 95%. Now its running for 2 hours without crashes.

1. The cards have Micron memory.
2. Done. Power management mode - Prefer maximum performance.
3. All of that I have done before I started mining.
4. I can turn on the fan speed, but the cards never got over 62C.

Nice!

So micron memory seems to be the main culprit.

Had some that could even take +700 memory but 8 out of 10 they tend to hit a stability wall around +500.

If you achieve stability at stock clocks then try to put +500mem max on one card and keep the rest stock, if it's stable move on and add +500 to the second etc
There maybe one unlucky sample that cannot afford that OC.

I have once encountered instability as low as +250 in Micron.

[EthereumBa]

Nice!

So micron memory seems to be the main culprit.

Had some that could even take +700 memory but 8 out of 10 they tend to hit a stability wall around +500.

If you achieve stability at stock clocks then try to put +500mem max on one card and keep the rest stock, if it's stable move on and add +500 to the second etc
There maybe one unlucky sample that cannot afford that OC.

I have once encountered instability as low as +250 in Micron.

I will try with an second PSU for sure. Because the thing that bothers me is that they worked for 3 days fully stable without any errors on +100 core clock, +650 memory clock and 75% power. With 187MH/s average. What changed suddenly, I really don't know.

We'll see how long will they work now with this settings. It's 2:30 hours for now, and counting.

And can it be that the MH/s jumped up when I changed the nVidia Control panel? Claymore reports 181.8MH/s on the run.

Thanks for the help and tips

EDIT: I'm not with my mining rig at the moment. But in a month or so I will do a detailed optimization. Separate clocks for every card to take out the maximum from the rig.

[MikeGR7]

Ofc the setting helped 

[Sgubela]

Should I set clocks and voltage within claymore miner itself or is MSI AB better? Thanks

[Gagana]

Hi, Mr.Claymore

Please, make the "-minspeed" option understand, that there is no internet connection, so it doesn't actually mine. So, the reason of hitting "-minspeed" limit is not a hardware issue. So, I don't want the miner restarts (reboot the PC) in this case. You understand, that it triggers all other signaling like e-mail/messages "I have to reboot", "I have been rebooted" and so on from every miner affected.

I guess that there are people who need the opposite from what you describe (they need the -minspeed to get triggered from internet connection loss). I am not one of them but I am sure that there are cases where that makes sense.

I agree with you though on what you are saying. I would modify your request like this:

"Please add a flag that indicates whether internet connection loss will trigger -minspeed"

I think there is another uncovered case that bothers. There is a set of cool "failover" options in the miner. But all of them don't handle the trivial "internet connection lost" problem. O would prefer to run a "no_internet.bat" instead of stupid switching between mining pools. So, I understand your point, but I think it's really different. But lets Mr.Claymore to decide.

[Gagana]

Hi, Mr.Claymore
Please, make the "-minspeed" option understand, that there is no internet connection, so it doesn't actually mine. So, the reason of hitting "-minspeed" limit is not a hardware issue. So, I don't want the miner restarts (reboot the PC) in this case. You understand, that it triggers all other signaling like e-mail/messages "I have to reboot", "I have been rebooted" and so on from every miner affected.

Ok I will add an option in next update so you can set if "-minspeed" should work if miner cannot reach pools or not.

Thanks a lot! Do you have ETA for the next update?

[Gagana]

Hi, Mr.Claymore
Please, make the "-minspeed" option understand, that there is no internet connection, so it doesn't actually mine. So, the reason of hitting "-minspeed" limit is not a hardware issue. So, I don't want the miner restarts (reboot the PC) in this case. You understand, that it triggers all other signaling like e-mail/messages "I have to reboot", "I have been rebooted" and so on from every miner affected.

I guess that there are people who need the opposite from what you describe (they need the -minspeed to get triggered from internet connection loss). I am not one of them but I am sure that there are cases where that makes sense.
I agree with you though on what you are saying. I would modify your request like this:
"Please add a flag that indicates whether internet connection loss will trigger -minspeed"

Ok I will add an option in next update so you can set if "-minspeed" should work if miner cannot reach pools or not.

Yes, the default behavior is exactly how i use this flag.

I am one of the users that suffer from the already reported problem:
Every 2- 3 days my internet connection drops and reconnects after 5 minutes.
After reconnection the cards work at a lower speed than normal almost half speed.
I can easily reproduce it with just unplugging and replugging the Ethernet cable.
Only solution was for me to manually close the miner and reopen it to regain full speed.

"-minspeed" option provided me a solution and when the internet reconnected, the miner detects the lower speed and restarts by itself and regains full speed.

If Claymore needs it, i can provide a video clip with this issue or whatever info he needs.

Thanks for development efforts either way!

Hmmm... I will try to check if the speed drop is connected with the "internet connection lost" problem... Thanks for pointing this out!

[golden0707]

For the ones who are complaining about force restarts and freeze i will tell you my experience, i had the same problems and maybe this will help you to:
I have 5 sapphire rx580 8GB and 1 msi rx580 gaming edition 8GB, windows 10 64bit ver 1703
1. Until robin hood drivers i didn't have any problems with claymore or drivers or freeze. After installing this drivers i could not overclock my cards from claymore so i needed to tweak them with Afterburner and BIOS edits, but again i could stil function properly
2. When the official drivers from amd (still beta, but from official site) where launched i installed them and every 2-3 days I needed to reboot my sister. No matter what i have tried: OC tweets (even with the same settings which previously worked), I`ve seen the spikes in energy drawn by the sistem like Claymore said is causing the freeze.
3. Tha final step to create a stable sistem:
 - uninstall drivers with AMD wizard which installs also the 17.11.1 drivers. I didn't need to uninstall them from safe mode
 - patch the drivers
 - uninstall Afterburner
 - installed Claymore 10.1 with same Config file which worked with Robin Hood drivers
 - set every card in Compute mode, reboot the sistem and everything was set.

I already have a week without freeze or reboots.
Claymore could set everything except the fan speed of my MSI card: cclock, mclock, cvddc, mvddc
With 17.11.1 drivers i observe also a 50-60w economy on my sistem consumed

I hope this will help also others!

Hi! Well, i think im doing something wrong .
Ive been using AB ( 4.4) to overclock and undervolt GPUs koz i couldnt undervolt it in CM. If in CM i set -cvddc lets say 980 , ok its about 975 ( says GPUZ) , set to 900 , its 990, set to 930 , its 985 .
I thought its is fixed in CM 10 + 17.11.1 combo . But not, i cant set desired voltage via CM.

What am I doing wrong ? 

[PeaMine]

I had to reinstall all drivers, and it seems only 2 of the 5 cards are showing up.
Mixture of 480s and 580s, 2 of 3 580s show up, 0 of 2 480s show up.
Tried so many driver installs.

I also have mixed amd card models, what i do to reinstall drivers is:

1 - amdcleanuputility to remove the old driver
2- install the new driver i want
3- use the atikmdag-patcher right after drive instalation before restarting

hope it helps

Just curious, do you have modded BIOS?
Pixel Clock Patcher I believe is only used to get a modded card working.
All of my cards are stock.