Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v15.0 (Windows/Linux) - page 1051.

fittsy

sr. member

Activity: 441

Merit: 252

Quote from: monsanto on November 22, 2016, 05:27:27 PM

Wow, very impressed with this miner. I haven't tried dual mining before because my rig is unstable with most miners but it works with this one Wink

.

I'm just wondering, if we are getting a certain hashrate mining Decred, should we expect to get a similar hashrate with Siacoin or Lbry?

No they are different algorithms and have different hashrates.

Best to try it with your rig and see. You can also set the intensity with -dcri which can change your hashrate.

monsanto

legendary

Activity: 1241

Merit: 1005

..like bright metal on a sullen ground.

Wow, very impressed with this miner. I haven't tried dual mining before because my rig is unstable with most miners but it works with this one Wink

.

I'm just wondering, if we are getting a certain hashrate mining Decred, should we expect to get a similar hashrate with Siacoin or Lbry?

mettalmag

legendary

Activity: 1084

Merit: 1003

≡v≡

Quote from: Dark-m-u on November 22, 2016, 05:04:57 PM

please help
I get a blue screen with all the modified roms

rx480 nitro+ oc and rx470 nitro+ oc

win 10
h81 pro btc

I guess you used wrong bios

Dark-m-u

newbie

Activity: 23

Merit: 0

please help
I get a blue screen with all the modified roms

rx480 nitro+ oc and rx470 nitro+ oc

win 10
h81 pro btc

AriesIV10

legendary

Activity: 1260

Merit: 1006

Mine for a Bit

Quote from: Claymore on November 22, 2016, 03:44:47 PM

Readme or OP, section "KNOWN ISSUES":

- AMD cards: GPU indexes in temperature control sometimes don't match GPU indexes in mining. Miner has to enumerate GPUs via OpenCL API to execute OpenCL code, and also it has to enumerate GPUs via ADL API to manage temperature/clock.
And order of GPUs in these lists can be different. There is no way to fix GPUs order automatically, thanks to AMD devs.
But you can do it manually. For example, if you have two cards, you can change their order by adding "-di 10". Another example, reverse order for six cards: "-di 543210".
Also you can do it automatically (experimental feature) with "-di detect" option.

PS. You can try to ask AMD devs somehow why their OpenCL implementation enumerates GPUs in such strange way. I'd like to know it too.

Thank You VERY MUCH! This will help out immensely.

Claymore

donator

Activity: 1610

Merit: 1325

Miners developer

Readme or OP, section "KNOWN ISSUES":

- AMD cards: GPU indexes in temperature control sometimes don't match GPU indexes in mining. Miner has to enumerate GPUs via OpenCL API to execute OpenCL code, and also it has to enumerate GPUs via ADL API to manage temperature/clock.
And order of GPUs in these lists can be different. There is no way to fix GPUs order automatically, thanks to AMD devs.
But you can do it manually. For example, if you have two cards, you can change their order by adding "-di 10". Another example, reverse order for six cards: "-di 543210".
Also you can do it automatically (experimental feature) with "-di detect" option.

PS. You can try to ask AMD devs somehow why their OpenCL implementation enumerates GPUs in such strange way. I'd like to know it too.

AriesIV10

legendary

Activity: 1260

Merit: 1006

Mine for a Bit

Quote from: Trimegistus on November 22, 2016, 12:13:46 PM

Quote from: Marcinprv on November 22, 2016, 11:43:09 AM

This is a NIGHTMARE! Angry

Do you have the same problem as me?

Yes...

I am with you on this one!!!!

By the way, very nice pictorial.

matmator

sr. member

Activity: 487

Merit: 252

bou !

lol

Trimegistus

legendary

Activity: 1564

Merit: 1027

Quote from: Marcinprv on November 22, 2016, 11:43:09 AM

This is a NIGHTMARE! Angry

Do you have the same problem as me?

Yes...

Marcinprv

newbie

Activity: 29

Merit: 0

This is a NIGHTMARE! Angry

https://s14.postimg.org/8feuzoqtt/Connection_Scheme.png

Do you have the same problem as me?

Correction - H81, not G81 Wink

not.you

legendary

Activity: 1726

Merit: 1018

Quote from: CoRpO on November 22, 2016, 10:41:04 AM

Quote from: not.you on November 22, 2016, 10:37:25 AM

Quote from: mettalmag on November 22, 2016, 09:31:07 AM

Quote from: not.you on November 22, 2016, 09:20:31 AM

I am having trouble getting my rx 480 rig to work for anything. I have tried multiple crimson drivers (I still suspect a driver problem) but I always get the same result. Claymore's miner crashes as soon as it would normally start hashing, and this is true for both this dual miner and Claymore's ZEC miner. Ethminer will run, but shows zero hashes (this is true for multiple versions of ethminer). I have tried removing and reinstalling 5 different crimson driver versions and it all seems to succeed successfully but I am still getting these same results every time. This PC was running catalyst 15.21 originally with r9 290's and I suspect some artifact from those drivers is sticking around. Before I proceed with an OS reinstall I wonder if anyone has suggestions. I have used DDU between crimson driver versions (although not in safe mode just yet). I am working remotely so can't do DDU in safe mode but I will be going onsite today to try that as my last resort before an OS reinstall. Anyone have a suggestion? (Win 7 x64 if it matters).

I suppose there are three cards in your rig, plugging in just one card to be sure PSU is ok would be a great idea before you start messing around with OS, when you get to that place you should try that first I think. If it's a driver stuff then reinstalling OS will fix it.

I have 2 GPU's. I don't think it is a PSU issue, the PSU is way overpowered for this and it worked fine with the R9 290's which theoretically should draw more power than the 480's. I am going to try it with just one GPU when I get there in a little while. I will also try adding some system RAM. It only has 2GB, I have slots and dimms to bring it up to 8 so I will do that.

IIRC you don't have enough ram for the rx4x0, they need more than the previous amd cards (even if they don't use it ... go figure)

I upgraded the RAM to 8GB and ran DDU in safe mode and one of the two seems to have made the difference because I am working on ETH and ZEC now.

CoRpO

full member

Activity: 185

Merit: 100

Quote from: not.you on November 22, 2016, 10:37:25 AM

Quote from: mettalmag on November 22, 2016, 09:31:07 AM

Quote from: not.you on November 22, 2016, 09:20:31 AM

I am having trouble getting my rx 480 rig to work for anything. I have tried multiple crimson drivers (I still suspect a driver problem) but I always get the same result. Claymore's miner crashes as soon as it would normally start hashing, and this is true for both this dual miner and Claymore's ZEC miner. Ethminer will run, but shows zero hashes (this is true for multiple versions of ethminer). I have tried removing and reinstalling 5 different crimson driver versions and it all seems to succeed successfully but I am still getting these same results every time. This PC was running catalyst 15.21 originally with r9 290's and I suspect some artifact from those drivers is sticking around. Before I proceed with an OS reinstall I wonder if anyone has suggestions. I have used DDU between crimson driver versions (although not in safe mode just yet). I am working remotely so can't do DDU in safe mode but I will be going onsite today to try that as my last resort before an OS reinstall. Anyone have a suggestion? (Win 7 x64 if it matters).

I suppose there are three cards in your rig, plugging in just one card to be sure PSU is ok would be a great idea before you start messing around with OS, when you get to that place you should try that first I think. If it's a driver stuff then reinstalling OS will fix it.

I have 2 GPU's. I don't think it is a PSU issue, the PSU is way overpowered for this and it worked fine with the R9 290's which theoretically should draw more power than the 480's. I am going to try it with just one GPU when I get there in a little while. I will also try adding some system RAM. It only has 2GB, I have slots and dimms to bring it up to 8 so I will do that.

IIRC you don't have enough ram for the rx4x0, they need more than the previous amd cards (even if they don't use it ... go figure)

not.you

legendary

Activity: 1726

Merit: 1018

Quote from: mettalmag on November 22, 2016, 09:31:07 AM

Quote from: not.you on November 22, 2016, 09:20:31 AM

I am having trouble getting my rx 480 rig to work for anything. I have tried multiple crimson drivers (I still suspect a driver problem) but I always get the same result. Claymore's miner crashes as soon as it would normally start hashing, and this is true for both this dual miner and Claymore's ZEC miner. Ethminer will run, but shows zero hashes (this is true for multiple versions of ethminer). I have tried removing and reinstalling 5 different crimson driver versions and it all seems to succeed successfully but I am still getting these same results every time. This PC was running catalyst 15.21 originally with r9 290's and I suspect some artifact from those drivers is sticking around. Before I proceed with an OS reinstall I wonder if anyone has suggestions. I have used DDU between crimson driver versions (although not in safe mode just yet). I am working remotely so can't do DDU in safe mode but I will be going onsite today to try that as my last resort before an OS reinstall. Anyone have a suggestion? (Win 7 x64 if it matters).

I suppose there are three cards in your rig, plugging in just one card to be sure PSU is ok would be a great idea before you start messing around with OS, when you get to that place you should try that first I think. If it's a driver stuff then reinstalling OS will fix it.

I have 2 GPU's. I don't think it is a PSU issue, the PSU is way overpowered for this and it worked fine with the R9 290's which theoretically should draw more power than the 480's. I am going to try it with just one GPU when I get there in a little while. I will also try adding some system RAM. It only has 2GB, I have slots and dimms to bring it up to 8 so I will do that.

CoRpO

full member

Activity: 185

Merit: 100

Quote from: mettalmag on November 22, 2016, 09:55:36 AM

Quote from: CoRpO on November 22, 2016, 09:49:08 AM

Quote from: mettalmag on November 22, 2016, 09:46:38 AM

Quote from: CoRpO on November 22, 2016, 09:39:21 AM

Quote from: mettalmag on November 22, 2016, 09:02:23 AM

I know that one dummy per rig and that's the strange thing, rigs were are 3x RX480 cards don't behave like that, rigs were are 6x RX480 cards randomly do this not all and not every time. Like if I restart rig were no monitor is plugged and it auto starts the miner in remote manager and even after connecting to RDP shows only one card stats, if I restart it second time it will work. For now two out of 13 6x RX480 rigs ar showing only one card stats. Actually I didn't mind it showing me less info, I just want to be secure in case of some card loses it's fan or any other reason overheats it, I just need the command -tstop to work. So the question will be if the miner sees the one card stats only will the command -tstop make other overheated card to stop ?

tstop won't work as the driver ins't giving any monitoring data to the miner
I tried everything in the registry (disablling ulps), or various adapters, nothing worked except a real monitor (even off) or a dummy plug. As soon as that was plugged, monitoring was back and I was able to access wattman settings

too bad, It costs damn too much. Buying two or three won't make any difference

Well, spending $12 on $750 rigs doesn't seem unreasonnable.
As it allows you to undervolt and to avoid 30% on your energy bill ...

Edit: It's even cheaper : http://www.ebay.com/itm/HDMI-dummy-plug-4K-EDID-emulator-monitor-for-mining-rigs-and-other-/232147267710?hash=item360d0e307e:g:2bkAAOSwcLxYEf2Z

Yeas I know but I have to test them first, buy two or three of those, there are 77 rigs total, some of them have other problems which involves plugged monitor so..

Testing part is pretty straightforward : works like if a monitor was plugged
I know that it seems expensive, that's what I though at first. Calculations been made, they where paid in less than a day ...

mettalmag

legendary

Activity: 1084

Merit: 1003

≡v≡

Quote from: CoRpO on November 22, 2016, 09:49:08 AM

Quote from: mettalmag on November 22, 2016, 09:46:38 AM

Quote from: CoRpO on November 22, 2016, 09:39:21 AM

Quote from: mettalmag on November 22, 2016, 09:02:23 AM

I know that one dummy per rig and that's the strange thing, rigs were are 3x RX480 cards don't behave like that, rigs were are 6x RX480 cards randomly do this not all and not every time. Like if I restart rig were no monitor is plugged and it auto starts the miner in remote manager and even after connecting to RDP shows only one card stats, if I restart it second time it will work. For now two out of 13 6x RX480 rigs ar showing only one card stats. Actually I didn't mind it showing me less info, I just want to be secure in case of some card loses it's fan or any other reason overheats it, I just need the command -tstop to work. So the question will be if the miner sees the one card stats only will the command -tstop make other overheated card to stop ?

tstop won't work as the driver ins't giving any monitoring data to the miner
I tried everything in the registry (disablling ulps), or various adapters, nothing worked except a real monitor (even off) or a dummy plug. As soon as that was plugged, monitoring was back and I was able to access wattman settings

too bad, It costs damn too much. Buying two or three won't make any difference

Well, spending $12 on $750 rigs doesn't seem unreasonnable.
As it allows you to undervolt and to avoid 30% on your energy bill ...

Edit: It's even cheaper : http://www.ebay.com/itm/HDMI-dummy-plug-4K-EDID-emulator-monitor-for-mining-rigs-and-other-/232147267710?hash=item360d0e307e:g:2bkAAOSwcLxYEf2Z

Yeas I know but I have to test them first, buy two or three of those, there are 77 rigs total, some of them have other problems which involves plugged monitor so..

CoRpO

full member

Activity: 185

Merit: 100

Quote from: mettalmag on November 22, 2016, 09:46:38 AM

Quote from: CoRpO on November 22, 2016, 09:39:21 AM

Quote from: mettalmag on November 22, 2016, 09:02:23 AM

I know that one dummy per rig and that's the strange thing, rigs were are 3x RX480 cards don't behave like that, rigs were are 6x RX480 cards randomly do this not all and not every time. Like if I restart rig were no monitor is plugged and it auto starts the miner in remote manager and even after connecting to RDP shows only one card stats, if I restart it second time it will work. For now two out of 13 6x RX480 rigs ar showing only one card stats. Actually I didn't mind it showing me less info, I just want to be secure in case of some card loses it's fan or any other reason overheats it, I just need the command -tstop to work. So the question will be if the miner sees the one card stats only will the command -tstop make other overheated card to stop ?

tstop won't work as the driver ins't giving any monitoring data to the miner
I tried everything in the registry (disablling ulps), or various adapters, nothing worked except a real monitor (even off) or a dummy plug. As soon as that was plugged, monitoring was back and I was able to access wattman settings

too bad, It costs damn too much. Buying two or three won't make any difference

Well, spending $12 on $750 rigs doesn't seem unreasonnable.
As it allows you to undervolt and to avoid 30% on your energy bill ...

Edit: It's even cheaper : http://www.ebay.com/itm/HDMI-dummy-plug-4K-EDID-emulator-monitor-for-mining-rigs-and-other-/232147267710?hash=item360d0e307e:g:2bkAAOSwcLxYEf2Z

Nickola

newbie

Activity: 8

Merit: 0

Quote from: Claymore on November 22, 2016, 09:34:36 AM

2. It seems font-family: monospace; does not work in tag, did not check it deeply though.

1)

or

2) before tag add:

mettalmag

legendary

Activity: 1084

Merit: 1003

≡v≡

Quote from: CoRpO on November 22, 2016, 09:39:21 AM

Quote from: mettalmag on November 22, 2016, 09:02:23 AM

I know that one dummy per rig and that's the strange thing, rigs were are 3x RX480 cards don't behave like that, rigs were are 6x RX480 cards randomly do this not all and not every time. Like if I restart rig were no monitor is plugged and it auto starts the miner in remote manager and even after connecting to RDP shows only one card stats, if I restart it second time it will work. For now two out of 13 6x RX480 rigs ar showing only one card stats. Actually I didn't mind it showing me less info, I just want to be secure in case of some card loses it's fan or any other reason overheats it, I just need the command -tstop to work. So the question will be if the miner sees the one card stats only will the command -tstop make other overheated card to stop ?

tstop won't work as the driver ins't giving any monitoring data to the miner
I tried everything in the registry (disablling ulps), or various adapters, nothing worked except a real monitor (even off) or a dummy plug. As soon as that was plugged, monitoring was back and I was able to access wattman settings

too bad, It costs damn too much. Buying two or three won't make any difference

CoRpO

full member

Activity: 185

Merit: 100

Quote from: mettalmag on November 22, 2016, 09:02:23 AM

I know that one dummy per rig and that's the strange thing, rigs were are 3x RX480 cards don't behave like that, rigs were are 6x RX480 cards randomly do this not all and not every time. Like if I restart rig were no monitor is plugged and it auto starts the miner in remote manager and even after connecting to RDP shows only one card stats, if I restart it second time it will work. For now two out of 13 6x RX480 rigs ar showing only one card stats. Actually I didn't mind it showing me less info, I just want to be secure in case of some card loses it's fan or any other reason overheats it, I just need the command -tstop to work. So the question will be if the miner sees the one card stats only will the command -tstop make other overheated card to stop ?

tstop won't work as the driver ins't giving any monitoring data to the miner
I tried everything in the registry (disablling ulps), or various adapters, nothing worked except a real monitor (even off) or a dummy plug. As soon as that was plugged, monitoring was back and I was able to access wattman settings

Claymore

donator

Activity: 1610

Merit: 1325

Miners developer

Quote from: Nickola on November 22, 2016, 08:58:36 AM

to Claymore:

Bug in EthMan 2.5 - "View miner console" command doesn't work at the first line of the list of miners.

Also, please add style element "font-family: monospace;" to tag "" in web-browser console.

1. Confirmed, I uploaded updated EthMan.exe to Dual-miner folder on MEGA.
2. It seems font-family: monospace; does not work in tag, did not check it deeply though.

Topic: Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v15.0 (Windows/Linux) - page 1051. (Read 6590565 times)