RTX 3080 error: unable to get fan speed and temperature - Unknown Error (15)

[murgorx]

Hello guys,

since yesterday I am experiecing a rather strange prboelm with my mining rig.

I stopped to blow and clean one of the cards and change the termal pads on it and after I plugged it back into the rig, one of the rest cards started crashing.

The rig is stable for a couple of minutes, then all of a sudden, one card just crashes and the phoenixminer shows the following error :


hwmc GPU1: unable to get temperature - GPU is lost (15)
hwmc GPU1: unable to get fan speed - GPU is lost (15)

The card in HWinfo is still shown, but in a bit more light-gray-ish colour, as if its grayed out and not in black, like the rest of the GPUs.

I am currently using Phoenixminer 5.5c and the latest drivers of nVidia(this was happening with the old ones too)

[1714786413]

1714786413

[1714786413]

1714786413

[rdluffy]

Did you change the position of the cards after the maintenance?

Try to unplug and plug everything again, checking every cable and PCIe connection
Uninstall drivers with DDU: https://www.guru3d.com/files-details/display-driver-uninstaller-download.html
Install drivers again

After that post here

But just a question, are you sure you didn't damage anything on your card when you opened?

[PerfectCircle]

Before cleaning the GPU you removed the card from a pcie x16 slot or a riser? If it's a riser make sure you resit the riser cable very well or maybe you probably get a defeat riser cable already? Already remove driver and reinstall again

[murgorx]

I have removed another card, not the one that's currently crashing.

The card was working without issues until now ( I changed its thermal pads 1 month ago and ever since then it was working super fine)
The riser is fine, at least not showing any problems. All cables are done right.

I have uninstalled my drivers with DDU and installed the newest ones via the GeForce experience.

I read in a forum that this may happen due to increased OC - so I lowered from +850mhz on the memory to +800 and will keep an eye on it for now.


EDIT:
I have lowered, as said, the overclock settings from 850mhs on the memory to a bit under 800 and for the moment its being stable for the past 11-12 minutes... 
I will keep watching and update you, if something new comes up.

BTW still looking for some info on what's causing this..

[murgorx]

Well, after about an hour of mining  the GPU had crashed with the same error, this time the miner continued mining with the rest 3 cards...
Trying with lower OC settings now and will replace the riser board in a few

[murgorx]

ROFL, I changed the riser board of the card that was crashing, swapped 2 GPUS and now the rig has gone totally mental.
Two other cards are failing, the rig crashes as soon as I start the miner... this is getting out of control.

I will maybe re-install windows, as when I restarted the rig the first time I took the card, it installed a "critical " upgrade, which took around 40 minutes and ever since then my Windows started acting like a fool..

[murgorx]

Seems like I am going to install a new Windows , this rig is getting out of control... I've tried to re-install the drivers with DDU, the rig goes black for 5 minutes, then boots up normally with nVidia GeForce Experience on it and the nVidia control panel Something feels fishy af with that Windows update I installed yesterday..

[rdluffy]

Something feels fishy af with that Windows update I installed yesterday..

After I changed Windows 10 to Hive Os I can rest without any worries, no more babysitting rig, here: https://hiveos.farm/
You just need an small SSD or usb flash stick and try it, it's way better than windows, my rig is stable and I only restart to clean, or any upgrade on hive, and it's just a couple of minutes
It's a life changer (serious )

You can try SimpleMining or another OS based on Linux

[murgorx]

Okay, I've rolled back to the previous version of MS Windows 10, which I also installed at the time i took the card to replace the thermal pads on it, and ever since the problems started.. since like this new "enhanced version " of windows is full of #$%^ and may cause such prbolems.

For now the rig is stable. I'll leave it for the night and check tomorrow. I will also provide screenshots to which Windows version NOT TO UPDATE

Peace.

@Luffy - I will most surely try HiveOS in the nearest future possible, I am just scared to make a mistake and overclock too much on it, haven't played with it yet and too comfy with Windows, but that was it I will switch all my rigs from Uncle Bill's rat poison to a Linux OS based one, fuck Windows and the comfort zone

[rdluffy]

Okay, I've rolled back to the previous version of MS Windows 10, which I also installed at the time i took the card to replace the thermal pads on it, and ever since the problems started.. since like this new "enhanced version " of windows is full of #$%^ and may cause such prbolems.

For now the rig is stable. I'll leave it for the night and check tomorrow. I will also provide screenshots to which Windows version NOT TO UPDATE

Peace.

@Luffy - I will most surely try HiveOS in the nearest future possible, I am just scared to make a mistake and overclock too much on it, haven't played with it yet and too comfy with Windows, but that was it I will switch all my rigs from Uncle Bill's rat poison to a Linux OS based one, fuck Windows and the comfort zone

It's nice to see your rig is probably fixed and stable again

I promise you, after try some mining OS you'll never want to use windows 10 again in your rigs, I only use Windows in my machine I use daily for general stuff, but my main rig is with HiveOs, running solid stable

About Windows updates I tried everything to stop, but after sometime all my settings were reseted and updates turned on again. You can try to find a tool to completely block updates, I know there's a lot of softwares doing that, and another useful link for you: https://1stminingrig.com/best-windows-setup-configuration-tweaks-for-mining/

You can find some great stuff about tweaking windows and even try to block updates

[Headsimon]

Guys i am having two GPUs, 3060ti Founder edition and 3060ti Gigabyte edition. Gigabyte one was older on which i have mined for 3 months with no issues. However after mounting it on riser, an error is coming up at random times. Sometimes after 2 hours, sometimes after 2 days. Error goes like this, "hwmc GPU2: unable to get GPU fan speed- GPU is lost (15)" and then it restarts the PC. Please help how to solve this.

[swogerino]

Glad that you fixed the problem,however remember that saying I have got 99 problems and one riser is all of them.Risers play a vital role now although the RTX 3080 as all other RTX 30 series are some of the most stable cards ever created in mining.Windows works just Linux,any Linux if you can fine tune it to your needs.Now the latest version of risers 010 is out so upgrading to this version can eliminate many of those 99 problems I said.

[FP91G]

Guys i am having two GPUs, 3060ti Founder edition and 3060ti Gigabyte edition. Gigabyte one was older on which i have mined for 3 months with no issues. However after mounting it on riser, an error is coming up at random times. Sometimes after 2 hours, sometimes after 2 days. Error goes like this, "hwmc GPU2: unable to get GPU fan speed- GPU is lost (15)" and then it restarts the PC. Please help how to solve this.

What kind of raiser do you use?
Such errors were in the 10X0 series video cards when they used the sata raisers. The raisers were in good working order, the problems were in the contacts with the power supply. These contacts became hot and the farm stopped working or rebooted.
You can also try to lower the temperature of the video cards by increasing the fan speed.

[Headsimon]

Guys i am having two GPUs, 3060ti Founder edition and 3060ti Gigabyte edition. Gigabyte one was older on which i have mined for 3 months with no issues. However after mounting it on riser, an error is coming up at random times. Sometimes after 2 hours, sometimes after 2 days. Error goes like this, "hwmc GPU2: unable to get GPU fan speed- GPU is lost (15)" and then it restarts the PC. Please help how to solve this.

What kind of raiser do you use?
Such errors were in the 10X0 series video cards when they used the sata raisers. The raisers were in good working order, the problems were in the contacts with the power supply. These contacts became hot and the farm stopped working or rebooted.
You can also try to lower the temperature of the video cards by increasing the fan speed.

The riser used by me is "https://www.amazon.in/PiPlusTM-VER009S-Powered-Adapter-Extension/dp/B08YXHPZXL/ref=sr_1_4?dchild=1&keywords=piplus+riser&qid=1632508986&sr=8-4". Please check. I am a newbie therefore dont know much technicalities. However i have powered riser and gpu from different PCIe cables from psu. I will try increasing fan speed but the issue is that it will run for a day or two nonstop and then suddenly in every 2 hours i get this error and miner reboots. Today it restarted the pc at 6:57, 8:57, 10:57 and so on. This is one log file for your reference,


2021.09.24:22:57:08.377: eths Eth: New job #efefc125 from ssl://asia1.ethermine.org:5555; diff: 4295MH
2021.09.24:22:57:10.022: hwmc GPU2: unable to get fan speed - GPU is lost (15)
2021.09.24:22:57:10.065: main GPU1: 68C 72% 140W, GPU2: 67C
GPUs power: 139.9 W; 428 kH/J
2021.09.24:22:57:11.320: main Eth speed: 59.923 MH/s, shares: 184/1/0, time: 1:58
2021.09.24:22:57:11.320: main GPUs: 1: 59.923 MH/s (90) 2: 0.000 MH/s (95)

2021.09.24:22:57:15.871: main *** 1:58 *** 9/24 22:57 **************************************
2021.09.24:22:57:15.871: main Eth: Mining ETH on ssl://asia1.ethermine.org:5555 for 1:16
2021.09.24:22:57:15.871: main Eth: Accepted shares 184 (1 stales), rejected shares 1 (0 stales)
2021.09.24:22:57:15.871: main Eth: Incorrect shares 0 (0.00%), est. stales percentage 0.54%
2021.09.24:22:57:15.871: main Eth: Maximum difficulty of found share: 2716.9 GH (!)
2021.09.24:22:57:15.871: main Eth: Average speed (5 min): 112.418 MH/s
2021.09.24:22:57:15.871: main Eth: Effective speed: 111.47 MH/s; at pool: 110.87 MH/s
2021.09.24:22:57:15.871: main  

after this mining is continued for 30 seconds and then it restarts the PC with windows failure of NO CARD DETECTED. Is the problem with riser? I tried changing the position of GPUs i.e. replaced two GPUs with one another and still the problem persists. Please help.

[swogerino]

Guys i am having two GPUs, 3060ti Founder edition and 3060ti Gigabyte edition. Gigabyte one was older on which i have mined for 3 months with no issues. However after mounting it on riser, an error is coming up at random times. Sometimes after 2 hours, sometimes after 2 days. Error goes like this, "hwmc GPU2: unable to get GPU fan speed- GPU is lost (15)" and then it restarts the PC. Please help how to solve this.

What kind of raiser do you use?
Such errors were in the 10X0 series video cards when they used the sata raisers. The raisers were in good working order, the problems were in the contacts with the power supply. These contacts became hot and the farm stopped working or rebooted.
You can also try to lower the temperature of the video cards by increasing the fan speed.

The riser used by me is "https://www.amazon.in/PiPlusTM-VER009S-Powered-Adapter-Extension/dp/B08YXHPZXL/ref=sr_1_4?dchild=1&keywords=piplus+riser&qid=1632508986&sr=8-4". Please check. I am a newbie therefore dont know much technicalities. However i have powered riser and gpu from different PCIe cables from psu. I will try increasing fan speed but the issue is that it will run for a day or two nonstop and then suddenly in every 2 hours i get this error and miner reboots. Today it restarted the pc at 6:57, 8:57, 10:57 and so on. This is one log file for your reference,


2021.09.24:22:57:08.377: eths Eth: New job #efefc125 from ssl://asia1.ethermine.org:5555; diff: 4295MH
2021.09.24:22:57:10.022: hwmc GPU2: unable to get fan speed - GPU is lost (15)
2021.09.24:22:57:10.065: main GPU1: 68C 72% 140W, GPU2: 67C
GPUs power: 139.9 W; 428 kH/J
2021.09.24:22:57:11.320: main Eth speed: 59.923 MH/s, shares: 184/1/0, time: 1:58
2021.09.24:22:57:11.320: main GPUs: 1: 59.923 MH/s (90) 2: 0.000 MH/s (95)

2021.09.24:22:57:15.871: main *** 1:58 *** 9/24 22:57 **************************************
2021.09.24:22:57:15.871: main Eth: Mining ETH on ssl://asia1.ethermine.org:5555 for 1:16
2021.09.24:22:57:15.871: main Eth: Accepted shares 184 (1 stales), rejected shares 1 (0 stales)
2021.09.24:22:57:15.871: main Eth: Incorrect shares 0 (0.00%), est. stales percentage 0.54%
2021.09.24:22:57:15.871: main Eth: Maximum difficulty of found share: 2716.9 GH (!)
2021.09.24:22:57:15.871: main Eth: Average speed (5 min): 112.418 MH/s
2021.09.24:22:57:15.871: main Eth: Effective speed: 111.47 MH/s; at pool: 110.87 MH/s
2021.09.24:22:57:15.871: main  

after this mining is continued for 30 seconds and then it restarts the PC with windows failure of NO CARD DETECTED. Is the problem with riser? I tried changing the position of GPUs i.e. replaced two GPUs with one another and still the problem persists. Please help.

That is a good riser and most probably you have problems with your overclock settings of the card.The correct settings for that card are 1-Power Limit 75%,2-Core Clock -200,3-Memory Clock +1200 max in Windows,if it doesn't support it and crashes go back to 1000 and see what happens,4-Fan settings some percentage which make it stay cool enough like less than 70C

Also to definitely remove the riser issue,change the USB cable of the riser with a new one or try if you can this newest version of the riser.

https://www.amazon.in/SaiRetail-Com-Express-Adapter-Capacitor-Mining/dp/B09FZ2PTJ3/ref=sr_1_1?dchild=1&keywords=riser+010s&qid=1632513876&sr=8-1

They added some troubleshooting lights which I assume are to make you easily identify if it is a riser problem,a PSU component or the GPU itself in this version.Always power up the riser with a 6 pin PCIE connector if you can.

[Headsimon]

Guys i am having two GPUs, 3060ti Founder edition and 3060ti Gigabyte edition. Gigabyte one was older on which i have mined for 3 months with no issues. However after mounting it on riser, an error is coming up at random times. Sometimes after 2 hours, sometimes after 2 days. Error goes like this, "hwmc GPU2: unable to get GPU fan speed- GPU is lost (15)" and then it restarts the PC. Please help how to solve this.

What kind of raiser do you use?
Such errors were in the 10X0 series video cards when they used the sata raisers. The raisers were in good working order, the problems were in the contacts with the power supply. These contacts became hot and the farm stopped working or rebooted.
You can also try to lower the temperature of the video cards by increasing the fan speed.

The riser used by me is "https://www.amazon.in/PiPlusTM-VER009S-Powered-Adapter-Extension/dp/B08YXHPZXL/ref=sr_1_4?dchild=1&keywords=piplus+riser&qid=1632508986&sr=8-4". Please check. I am a newbie therefore dont know much technicalities. However i have powered riser and gpu from different PCIe cables from psu. I will try increasing fan speed but the issue is that it will run for a day or two nonstop and then suddenly in every 2 hours i get this error and miner reboots. Today it restarted the pc at 6:57, 8:57, 10:57 and so on. This is one log file for your reference,


2021.09.24:22:57:08.377: eths Eth: New job #efefc125 from ssl://asia1.ethermine.org:5555; diff: 4295MH
2021.09.24:22:57:10.022: hwmc GPU2: unable to get fan speed - GPU is lost (15)
2021.09.24:22:57:10.065: main GPU1: 68C 72% 140W, GPU2: 67C
GPUs power: 139.9 W; 428 kH/J
2021.09.24:22:57:11.320: main Eth speed: 59.923 MH/s, shares: 184/1/0, time: 1:58
2021.09.24:22:57:11.320: main GPUs: 1: 59.923 MH/s (90) 2: 0.000 MH/s (95)

2021.09.24:22:57:15.871: main *** 1:58 *** 9/24 22:57 **************************************
2021.09.24:22:57:15.871: main Eth: Mining ETH on ssl://asia1.ethermine.org:5555 for 1:16
2021.09.24:22:57:15.871: main Eth: Accepted shares 184 (1 stales), rejected shares 1 (0 stales)
2021.09.24:22:57:15.871: main Eth: Incorrect shares 0 (0.00%), est. stales percentage 0.54%
2021.09.24:22:57:15.871: main Eth: Maximum difficulty of found share: 2716.9 GH (!)
2021.09.24:22:57:15.871: main Eth: Average speed (5 min): 112.418 MH/s
2021.09.24:22:57:15.871: main Eth: Effective speed: 111.47 MH/s; at pool: 110.87 MH/s
2021.09.24:22:57:15.871: main  

after this mining is continued for 30 seconds and then it restarts the PC with windows failure of NO CARD DETECTED. Is the problem with riser? I tried changing the position of GPUs i.e. replaced two GPUs with one another and still the problem persists. Please help.

That is a good riser and most probably you have problems with your overclock settings of the card.The correct settings for that card are 1-Power Limit 75%,2-Core Clock -200,3-Memory Clock +1200 max in Windows,if it doesn't support it and crashes go back to 1000 and see what happens,4-Fan settings some percentage which make it stay cool enough like less than 70C

Also to definitely remove the riser issue,change the USB cable of the riser with a new one or try if you can this newest version of the riser.

https://www.amazon.in/SaiRetail-Com-Express-Adapter-Capacitor-Mining/dp/B09FZ2PTJ3/ref=sr_1_1?dchild=1&keywords=riser+010s&qid=1632513876&sr=8-1

They added some troubleshooting lights which I assume are to make you easily identify if it is a riser problem,a PSU component or the GPU itself in this version.Always power up the riser with a 6 pin PCIE connector if you can.

My OC settings are Power limit 70, core clock tried a whole range of 0,-200,-300,-502, mem clock 900,950,1000,1100,1150,1200 fan speed such that it stays around 67 C. This riser is relatively new, should i go for it? In fact i ordered this yesterday, https://www.amazon.in/Tapia-V009S-Plus-Indicator-Extension/dp/B097R24JFB/ref=sr_1_5?dchild=1&keywords=riser&qid=1632540537&sr=8-5. Should i cancel it and reorder? This is the another 010s riser i could find https://www.vedantcomputers.com/riser-extender-ver0010s-plus-8-capacitor-ultra-stable-extra-led-16-into-to%20-1-into-power-pcie-80cm-usb-3point0-cable?search=riser.

[swogerino]

If you can get it fast shipped to you meaning you are buying locally then I would say it would be a good idea to try the new more robust 010S riser.If you buy internationally leave it as even the version 009S should do well.One thing you can do to try and see if the card has not problems,is remove all other cards and put this one alone in the PCIEX 16 slot of the motherboard and if it mines correctly for let's say 2 days then try it alone without any other card through every riser and see what happens,if the same problems happen with all the risers then this card needs the latest one but if it works well with some of the existing ones and only crashes with a few then you have a couple of bad risers.Try this to isolate your problem,I know it takes time and lost amount of money during the 2 days of testing but trust me in the end it will be worth it.

[Headsimon]

If you can get it fast shipped to you meaning you are buying locally then I would say it would be a good idea to try the new more robust 010S riser.If you buy internationally leave it as even the version 009S should do well.One thing you can do to try and see if the card has not problems,is remove all other cards and put this one alone in the PCIEX 16 slot of the motherboard and if it mines correctly for let's say 2 days then try it alone without any other card through every riser and see what happens,if the same problems happen with all the risers then this card needs the latest one but if it works well with some of the existing ones and only crashes with a few then you have a couple of bad risers.Try this to isolate your problem,I know it takes time and lost amount of money during the 2 days of testing but trust me in the end it will be worth it.

I get your point, I'll tell you what I did. The Gigabyte card let's say card A was continuously mining for 3 months. And GPU B when arrived, since I had no riser so plugged in GPU B and removed GPU A. This way I mined for 2 weeks. No problem so far. Riser arrives, I put card A in riser and B in mobo. After a week or two problem hits. Since then I've tried changing psu cables, changing 1x slot in Mobo for riser communication, swapping card A and B with each other. Problem has not been solved so far. Every time card with riser shows this error. Kindly comment whether my diagnosis can be concluded that riser has some issues.

As per your suggestion, I think I'll go for 010s riser and hope this problem gets solved.

[FP91G]

If you can get it fast shipped to you meaning you are buying locally then I would say it would be a good idea to try the new more robust 010S riser.If you buy internationally leave it as even the version 009S should do well.One thing you can do to try and see if the card has not problems,is remove all other cards and put this one alone in the PCIEX 16 slot of the motherboard and if it mines correctly for let's say 2 days then try it alone without any other card through every riser and see what happens,if the same problems happen with all the risers then this card needs the latest one but if it works well with some of the existing ones and only crashes with a few then you have a couple of bad risers.Try this to isolate your problem,I know it takes time and lost amount of money during the 2 days of testing but trust me in the end it will be worth it.

I get your point, I'll tell you what I did. The Gigabyte card let's say card A was continuously mining for 3 months. And GPU B when arrived, since I had no riser so plugged in GPU B and removed GPU A. This way I mined for 2 weeks. No problem so far. Riser arrives, I put card A in riser and B in mobo. After a week or two problem hits. Since then I've tried changing psu cables, changing 1x slot in Mobo for riser communication, swapping card A and B with each other. Problem has not been solved so far. Every time card with riser shows this error. Kindly comment whether my diagnosis can be concluded that riser has some issues.

As per your suggestion, I think I'll go for 010s riser and hope this problem gets solved.

I will tell you from my own experience that such problems arise either because of the power supply, or because of problems in contacts.
To check, you need to have other raisers and check the power supply cables. Also check the 12 volt voltage with a multimeter under load.
If the voltage is below 11.5 volts, then your power supply is bad.
I do not remember (did not find) the model of your power supply.

[Headsimon]

If you can get it fast shipped to you meaning you are buying locally then I would say it would be a good idea to try the new more robust 010S riser.If you buy internationally leave it as even the version 009S should do well.One thing you can do to try and see if the card has not problems,is remove all other cards and put this one alone in the PCIEX 16 slot of the motherboard and if it mines correctly for let's say 2 days then try it alone without any other card through every riser and see what happens,if the same problems happen with all the risers then this card needs the latest one but if it works well with some of the existing ones and only crashes with a few then you have a couple of bad risers.Try this to isolate your problem,I know it takes time and lost amount of money during the 2 days of testing but trust me in the end it will be worth it.

I get your point, I'll tell you what I did. The Gigabyte card let's say card A was continuously mining for 3 months. And GPU B when arrived, since I had no riser so plugged in GPU B and removed GPU A. This way I mined for 2 weeks. No problem so far. Riser arrives, I put card A in riser and B in mobo. After a week or two problem hits. Since then I've tried changing psu cables, changing 1x slot in Mobo for riser communication, swapping card A and B with each other. Problem has not been solved so far. Every time card with riser shows this error. Kindly comment whether my diagnosis can be concluded that riser has some issues.

As per your suggestion, I think I'll go for 010s riser and hope this problem gets solved.

I will tell you from my own experience that such problems arise either because of the power supply, or because of problems in contacts.
To check, you need to have other raisers and check the power supply cables. Also check the 12 volt voltage with a multimeter under load.
If the voltage is below 11.5 volts, then your power supply is bad.
I do not remember (did not find) the model of your power supply.

Thanks. I have ordered a new riser. Meanwhile this is my psu link https://www.antec.com/product/power/hcg-gold850