c:\windows\system32\nvml.dll
C:\Program Files\NVIDIA Corporation\NVSMI\nvml.dll
C:\Program Files (x86)\NVIDIA Corporation\NVSMI\nvml.dll
So I don't think it's a problem with AwesomeMiner after all.
I just spotted this on on the command prompt window for PhoenixMiner the rig while doing the search.
GPU1: N/A, GPU2: N/A, GPU3: 59C 53% 115W, GPU4: N/A, GPU5: N/A, GPU6: 59C 53% 135W, GPU7: 58C 52% 137W, GPU8: N/A, GPU9: N/A, GPU10: N/A, GPU11: N/A
GPUs power: 387.0 W
I tried switching to Claymore and it yielded the same result in AwesomeMiner.
Strange, isn't it?
Thanks for sharing all details - it was really helpful. I have an idea what it's about and there will soon be an updated release with some improvements. As I'm not able to reproduce this in my setup, please let me know the result once the new version is available. Thanks! Wow! I'm not sure what you did but it seemed to have worked. Thanks! Good thing too because the 1600W PDU that I had ordered requires a C19 cable (instead of the expected C13) to connect it to my 220V PDU. So it looks like it might be a few more days until I get it set up.  Thanks for your diligence. Thanks for the feedback. What I noticed based on your reports was that the nVidia driver installer only put nvml.dll in the Windows System32 folder these days. In the past the file was not installed in System32 at all. The other folders (Program Files) had an older nvml.dll not updated and not removed by the later driver versions. So in some systems you could have an outdated nvml.dll here, causing problems with Power Usage reporting for many applications incl. Awesome Miner. I changed Awesome Miner to favor the version in the System32 folder and this resolved the problem. You're a genius. Thanks again. |
|
|
|
@Binary100100
Maybe that help you:
Create folder NVSMI in C:/Programm Files/NVIDIA Corporation
find on computer file nvml.dll and copy to NVSMI
restart miner
Thanks for trying, but no. I even tried re-installing the drivers and rebooting. Same results. Thanks for sharing the fact that the LibreHardwareMonitor tool was working. That tool uses NVML in a similar way as Awesome Miner, so Awesome Miner is expected to work in this environment as well. I will investigate the issue in more detail. I may get back to you about the log file for Remote Agent if needed. Okay, thanks for trying. I'm expecting to get my new 1600W PSU on Tuesday. Depending on the time of the delivery I might be installing it as late as Wednesday. At that point, I might be moving the 1070 cards back to their original machine. At that point, this won't really matter much to me anymore. I only brought it to your attention because it was a pretty strange issue. If possible, could you please search for "nvml.dll" and let me know the path or paths where it's found? Many thanks! c:\windows\system32\nvml.dll C:\Program Files\NVIDIA Corporation\NVSMI\nvml.dll C:\Program Files (x86)\NVIDIA Corporation\NVSMI\nvml.dll So I don't think it's a problem with AwesomeMiner after all. I just spotted this on on the command prompt window for PhoenixMiner the rig while doing the search. GPU1: N/A, GPU2: N/A, GPU3: 59C 53% 115W, GPU4: N/A, GPU5: N/A, GPU6: 59C 53% 135W, GPU7: 58C 52% 137W, GPU8: N/A, GPU9: N/A, GPU10: N/A, GPU11: N/A GPUs power: 387.0 W I tried switching to Claymore and it yielded the same result in AwesomeMiner. Strange, isn't it? Thanks for sharing all details - it was really helpful. I have an idea what it's about and there will soon be an updated release with some improvements. As I'm not able to reproduce this in my setup, please let me know the result once the new version is available. Thanks! Wow! I'm not sure what you did but it seemed to have worked. Thanks! Good thing too because the 1600W PDU that I had ordered requires a C19 cable (instead of the expected C13) to connect it to my 220V PDU. So it looks like it might be a few more days until I get it set up. Thanks for your diligence. |
|
|
|
@Binary100100
Maybe that help you:
Create folder NVSMI in C:/Programm Files/NVIDIA Corporation
find on computer file nvml.dll and copy to NVSMI
restart miner
Thanks for trying, but no. I even tried re-installing the drivers and rebooting. Same results. Thanks for sharing the fact that the LibreHardwareMonitor tool was working. That tool uses NVML in a similar way as Awesome Miner, so Awesome Miner is expected to work in this environment as well. I will investigate the issue in more detail. I may get back to you about the log file for Remote Agent if needed. Okay, thanks for trying. I'm expecting to get my new 1600W PSU on Tuesday. Depending on the time of the delivery I might be installing it as late as Wednesday. At that point, I might be moving the 1070 cards back to their original machine. At that point, this won't really matter much to me anymore. I only brought it to your attention because it was a pretty strange issue. If possible, could you please search for "nvml.dll" and let me know the path or paths where it's found? Many thanks! c:\windows\system32\nvml.dll C:\Program Files\NVIDIA Corporation\NVSMI\nvml.dll C:\Program Files (x86)\NVIDIA Corporation\NVSMI\nvml.dll So I don't think it's a problem with AwesomeMiner after all. I just spotted this on on the command prompt window for PhoenixMiner the rig while doing the search. GPU1: N/A, GPU2: N/A, GPU3: 59C 53% 115W, GPU4: N/A, GPU5: N/A, GPU6: 59C 53% 135W, GPU7: 58C 52% 137W, GPU8: N/A, GPU9: N/A, GPU10: N/A, GPU11: N/A GPUs power: 387.0 W I tried switching to Claymore and it yielded the same result in AwesomeMiner. Strange, isn't it? |
|
|
|
@Binary100100
Maybe that help you:
Create folder NVSMI in C:/Programm Files/NVIDIA Corporation
find on computer file nvml.dll and copy to NVSMI
restart miner
Thanks for trying, but no. I even tried re-installing the drivers and rebooting. Same results. Thanks for sharing the fact that the LibreHardwareMonitor tool was working. That tool uses NVML in a similar way as Awesome Miner, so Awesome Miner is expected to work in this environment as well. I will investigate the issue in more detail. I may get back to you about the log file for Remote Agent if needed. Okay, thanks for trying. I'm expecting to get my new 1600W PSU on Tuesday. Depending on the time of the delivery I might be installing it as late as Wednesday. At that point, I might be moving the 1070 cards back to their original machine. At that point, this won't really matter much to me anymore. I only brought it to your attention because it was a pretty strange issue. |
|
|
|
@Binary100100
Maybe that help you:
Create folder NVSMI in C:/Programm Files/NVIDIA Corporation
find on computer file nvml.dll and copy to NVSMI
restart miner
Thanks for trying, but no. I even tried re-installing the drivers and rebooting. Same results. |
|
|
|
Thanks for sharing all details.
For the power usage the nVidia drivers are indicating an error when Awesome Miner query for the power usage. In case you have more mining rigs, is this the only one where you don't get nVidia power usage? If you restart Awesome Miner and select the System tab at the bottom for this miner - do you see any power usage?
For the device ordering - in the dialog "Map to system monitoring" you can manually set the correct order by using the Up and Down arrows (in upper right corner). I do intent do improve how Awesome Miner automatically tries to sort out the ordering of the GPU's, but the difficulty is that many mining software report the hashrate for the GPU's in their own order - and Awesome Miner isn't yet smart enough to sort this out.
It shows everything except the power usage. Displays as 0 W. But in GPU-Z everything shows as it should. Suggestions? Hi, Thanks for the update. A few questions: 1) Does a reboot of the Remote Agent computer resolve the issue? 2) Is this problem only present on this specific mining rig or on other as well? 3) GPUz has some special sensor reading capabilities and isn't using nVidia NVML as Awesome Miner and most other software do. If you run some mining software that display power usage in the mining console window - is that one correctly displaying the power usage on this computer? 1) I shut down my rigs every night. Makes sleeping easier.  So no. 2) I didn't have this problem with the original machine (with the bad PSU) that I took these 8 cards from. 3) It shows the wattage on LibreHardwareMonitor:  |
|
|
|
So I had to transfer 8 of my NVIDIA 1070 cards to my (WIP) rig that already had 3 RX 5700 cards on it because I was having a problem with the PSU on the 1070 rig.
But after I did, everything got all messed up on the GPU tab and it won't display watts for my NVIDIA cards.
As you'll see GPU 0 and 1 (5700 XT) is reading the hashrate of a 1070
GPU 2 (5700 XT) is reading for a 5700 XT (but the right one?)
GPU 5 and 6 are 1070 cards but read the hashrate of my 5700 XT cards
None of the cards 3 - 10 are displaying the wattage.
Advice?
Could you please click the "View GPU Details" button in this GPU tab and send me the output? Please send via PM or mail because of the size. Thanks! To some extent the "Map to system monitoring" button can be used to reorder the GPU's, but that will not resolve the missing power display. Check your messages. Thanks. The "map to system monitoring" seems to have something crossed though. My RX 5700 XT cards are 1780 MHz Memory and the 1070 cards are 4428 MHz. Significant difference. But where is posts the "Performance" is where it's interesting. GPU 0 is displayed as hashing ETH at 31.42MH/s and GPU 1 is 30.53. Both are 5700 XT cards. But GPU 2 is a 5700 XT card and hashes at 50.64 MH/s. GPU 3 is a 1070 which reads properly at 31.66 MH/s But GPU 5 is a 1070 which is reading at 50.8 MH/s. (Which is impossible. It's obviously a 5700 XT card.) Same with GPU 6. Regarding the "Map Device Order" I'm thinking that the problem is from the PCI Bus ID: The order is PCI Bus ID: 11 PCI Bus ID: 5 PCI Bus ID: 14 PCI Bus ID: 1 PCI Bus ID: 2 PCI Bus ID: 6 PCI Bus ID: 8 ... etc. Is that normal? Thanks for sharing all details. For the power usage the nVidia drivers are indicating an error when Awesome Miner query for the power usage. In case you have more mining rigs, is this the only one where you don't get nVidia power usage? If you restart Awesome Miner and select the System tab at the bottom for this miner - do you see any power usage? For the device ordering - in the dialog "Map to system monitoring" you can manually set the correct order by using the Up and Down arrows (in upper right corner). I do intent do improve how Awesome Miner automatically tries to sort out the ordering of the GPU's, but the difficulty is that many mining software report the hashrate for the GPU's in their own order - and Awesome Miner isn't yet smart enough to sort this out. It shows everything except the power usage. Displays as 0 W. But in GPU-Z everything shows as it should.  Suggestions?  |
|
|
|
So I had to transfer 8 of my NVIDIA 1070 cards to my (WIP) rig that already had 3 RX 5700 cards on it because I was having a problem with the PSU on the 1070 rig.
But after I did, everything got all messed up on the GPU tab and it won't display watts for my NVIDIA cards.
As you'll see GPU 0 and 1 (5700 XT) is reading the hashrate of a 1070
GPU 2 (5700 XT) is reading for a 5700 XT (but the right one?)
GPU 5 and 6 are 1070 cards but read the hashrate of my 5700 XT cards
None of the cards 3 - 10 are displaying the wattage.
Advice?
Could you please click the "View GPU Details" button in this GPU tab and send me the output? Please send via PM or mail because of the size. Thanks! To some extent the "Map to system monitoring" button can be used to reorder the GPU's, but that will not resolve the missing power display. Check your messages. Thanks. The "map to system monitoring" seems to have something crossed though. My RX 5700 XT cards are 1780 MHz Memory and the 1070 cards are 4428 MHz. Significant difference. But where is posts the "Performance" is where it's interesting. GPU 0 is displayed as hashing ETH at 31.42MH/s and GPU 1 is 30.53. Both are 5700 XT cards. But GPU 2 is a 5700 XT card and hashes at 50.64 MH/s. GPU 3 is a 1070 which reads properly at 31.66 MH/s But GPU 5 is a 1070 which is reading at 50.8 MH/s. (Which is impossible. It's obviously a 5700 XT card.) Same with GPU 6. Regarding the "Map Device Order" I'm thinking that the problem is from the PCI Bus ID: The order is PCI Bus ID: 11 PCI Bus ID: 5 PCI Bus ID: 14 PCI Bus ID: 1 PCI Bus ID: 2 PCI Bus ID: 6 PCI Bus ID: 8 ... etc. Is that normal?  |
|
|
|
So I had to transfer 8 of my NVIDIA 1070 cards to my (WIP) rig that already had 3 RX 5700 cards on it because I was having a problem with the PSU on the 1070 rig. But after I did, everything got all messed up on the GPU tab and it won't display watts for my NVIDIA cards. As you'll see GPU 0 and 1 (5700 XT) is reading the hashrate of a 1070 GPU 2 (5700 XT) is reading for a 5700 XT (but the right one?) GPU 5 and 6 are 1070 cards but read the hashrate of my 5700 XT cards None of the cards 3 - 10 are displaying the wattage. Advice?  |
|
|
|
I haven't posted in this thread for a while (sorry) but I was just playing around with "Balance" and attempted to set the address for my ETC address on my Ledger. Unforuntately, it failed. It appears that "Gastracker is shooting down its services" (ref: https://bye.gastracker.io/?page=) What are our alternatives for the block explorer API URL and can a fix be implemented in a near update? I've found an another source for ETC balance, but it returns the data in a format that required some changes to Awesome Miner. The next version of Awesome Miner will support the following - but it will not work with the current version https://blockscout.com/etc/mainnet/api?module=account&action=balance&address=[address] Awesome (pun intended) thanks for your continuous support. |
|
|
|
|
I'd suggest making sure you're cards were reset to Graphics Mode instead of Compute Mode.
Try adding -acm in your batch file or press "y" in the console.
|
|
|
|
I haven't posted in this thread for a while (sorry) but I was just playing around with "Balance" and attempted to set the address for my ETC address on my Ledger. Unforuntately, it failed. It appears that "Gastracker is shooting down its services" (ref: https://bye.gastracker.io/?page=) What are our alternatives for the block explorer API URL and can a fix be implemented in a near update? |
|
|
|
|
These are my settings:
-tt 58 -cvddc 750 -cclock 1300 -fanmin 0 -mclock 900 -amd -mode 1 -clKernel 1 -clNew 1
Anything higher than 900 mclock (even a 905) crashes after about 5 minutes.
I'm using the XFX 5700XT reference versions.
With these settings it's rock solid.
Between 58C to 60C and pulling between 93 - 95 watts.
|
|
|
|
Did you try increasing your virtual memory? I always heard that you should have a 1:1 for your combined card memory. So in your case I'd go with 64gb. On a 120GB hard drive that might be difficult. Personally I have 500GB SSD on each of my rigs. Maybe you can add another cheap SSD and just use that for VM?
You could also try just enabling one card at a time and see how many work?
Miner Properties > GPU Selection at the bottom > Use selected from list > (Select just one or two then hit OK) > See if it runs (stable.) If so, add another until it stops.
I also use Awesome Miner for all of my rigs and I'm not having any issues with Claymore 15. Alternatively, try PhoenixMiner. If you have the same problem then it's probably not with the software.
I got this sorted same day... Your solution was spot on, I have now set 40GB Virtual memory... worked a charm, much appreciated for the good info. Good to hear and thanks for the reply. |
|
|
|
6 x gigabyte GTX 1070 8gb
Windows 10
4GB DDR4 Ram on Motherboard
Additional 16-24GB Virtual Memory.
60GB SSD drive.
the farm stopped working on 6 cards, shows the DAG file in red on any card and immediately reboots the miner, disconnected one card, it works with 5 cards, please help.
Increase the virtual memory by 8gb. |
|
|
|
|
I'm worried that we're all going to be screwed if ETH/ETC goes POS... or even ProgPow. I must have have to sell my 5700 cards.
|
|
|
|
|
I wanted to build a 12 card rig with RX5700's but I'm worried about the future of the card because nobody knows for sure what's going to happen with ETH and ETC. (ProgPow and POS)
|
|
|
|
Have similar issue,
Updated my miners to Claymore 15 and now they just keep crashing...
8 x Asus RX 580 OC 8GB edition per mining rig
Windows 10 with Latest Update
8GB DDR4 Ram on Motherboard
Additional 24GB Virtual Memory.
120GB SSD drive.
Using older AMD ASUS Drivers that were working fine before update
Everything was working fine until new Claymore 15 released...hoping it is a bug and will be fixed soon.
Some of the GPUs start up fine but then the miner crashes and restarts
I am using latest version of Awesome Miner to run my mining farm. Anyone else having similar issues?
Any help would be much much appreciated.
Cheers
Did you try increasing your virtual memory? I always heard that you should have a 1:1 for your combined card memory. So in your case I'd go with 64gb. On a 120GB hard drive that might be difficult. Personally I have 500GB SSD on each of my rigs. Maybe you can add another cheap SSD and just use that for VM? You could also try just enabling one card at a time and see how many work? Miner Properties > GPU Selection at the bottom > Use selected from list > (Select just one or two then hit OK) > See if it runs (stable.) If so, add another until it stops. I also use Awesome Miner for all of my rigs and I'm not having any issues with Claymore 15. Alternatively, try PhoenixMiner. If you have the same problem then it's probably not with the software. |
|
|
|
Hey guys, i used to read last 50 pages before but i didnt find the solve. I have 5 rigs with 8x470/480/580 and all of them worked perfectly utill couple days ago... with claymore v10.. i was getting about 30hs from all of them and now when i tryied clayomor v15 only about 9hs.. im running it on win10, and used to overclock them with .bat file... Please can anyone help me, or show me the post where i can find the answers...
Sounds like you are using Graphics Mode instead of Compute Mode. |
|
|
|
I have made one custom windows especially for mining, I have removed almost all useles junk, including windows defender, windows update, store, xbox, services, programs, etc. I use it since 1 month and 0 errors, 0 restarts/updates/troubles. You can try it. Is original Windows 10 PRO 1709 ISO downloaded from microsoft, modded with N-Lite software tool version 1.8.0.6790. During installation will ask for serial (you can skip it as usual) and activation too after finishing. You can use any valid serial of Win 10 PRO (only pro), if you want to activate it. Cheers! https://drive.google.com/file/d/1kyxWZm0m17Of0MlE-wXjfng25yO7ZaAr/viewNot sure I trust a modded Windows 10 OS. No offense but what's to prevent you from tossing in something malicious? (Trojan, Keylogger, Packetsniffer etc.) I like the idea though. |
|
|
|
|
Show Posts
