Bitcoin Forum

So... I've been having the same problem for two weeks. I set up my miner, everything is running great, and then I launch Claymore, it loads for about 15 seconds and then the all hell breaks loose. The fans all turn to 100% and it sounds like a jet engine, then Claymore gives me and error about not being able to read temperature, and then my entire system crashes. Blue screen of "something went wrong. need to restart". Sometimes I get a warning saying "the application claymoreminer.... has been blocked from accessing graphics hardware". I have to reset the rig by holding down the power button. Sometimes when it loads back up I only have 3 graphics cards recognized and I basically have to reinstall windows because I cant get it to re-recognize the cards.

So far, here are a bunch of steps I've taken
PCIE lanes to Gen1
Disable CSM support
Virtual memory size to 24gb
Hi power mode on power settings and system settings
Reinstalled GPU drivers (8-10 times) directly from Nvidia
Turned off windows updates
Turned off windows defender (which is odd because it still blocks my miner when defender is off)
Placed my miner in an exception folder on windows defender
Switched to a different miner (ZCash excavator)
Reinstalled windows (4 times)
Reinstalled graphics cards 1 at a time
Purchased a new motherboard (thought it was the issue)
Turn off all overclocks
Updated motherboard bios

On both motherboards I have had the exact same issue. I think it's related to the nvidia drivers? Anyone PLEASE help.

My setup is
Gigabyte GA-270p-d3 motherboard
G3900 CPU
4gb DDR4
120gb m.2 SSD
x6 Asus gtx 1060 6gb
1000w rosewill power supply

I'm so frustrated I don't even know where to go next. Willing to send a few days worth of mining to whoever can figure this out. I'm desperate

enable 4g encoding in your bios if not already
also try gen2 in settings (my Z270A-P needed gen 2)
are you getting "system interrupts" in task manager ?
 

Interestingly, this motherboard doesn't have 4g encoding.

Tried gen2. No difference

Can you explain a bit more? not sure what system interrupts are.

Check if there is an bios update for you....

Just read your message.
"the application claymoreminer.... has been blocked from accessing graphics hardware"

This is not bios related...

That message is common with Vegas. Could be several things. Most common is Windows Defender.... add the file, the folder and if its already there, erase it and add it again. Even if its disable.
Do you have any other anti virus:

hmm, interesting, prob not 4g encoding if you boot all cards into windows successfully anyways actually now that i think about it :P

system interrupts are caused by hardware/driver errors, and they can eat up resources till your PC crashes
if you have a faulty riser, you can get them, but they wont start till you start mining .
you will see in task manager "system interrupts" with high cpu usage
solution for them is update drivers (system drivers not GPU)/replace faulty riser usually

Yeah before the crash I can get it to recognize all 6 cards. The problem is once I launch the miner

Ok... now we may be getting somewhere. I tried two tests.

First I launched it with all of the GPUs unplugged except one. On a single 1060 it seems to work just fine. I tested a few other GPUs by themselves and they also will mine perfectly... until a second GPU is plugged in.

Once I plugged in the second one, crash.

I had the task manager open and I noticed before it crashed the CPU load shot up to 80-90%. I also did see a program in there called "system interrupts" but it wasn't accounting for much of the CPU spike according to the task manager

ya man, test each card till error appears, replace riser on the faulty one may fix it (hopefully) most risers have a terrible fail rate, sometime right OOB, sometimes after a few weeks.months mining :(
i run also a free iobit softwear called "driver booster". free and works. it usually finds drivers for "PCI-E" stuff, and have had it help having all those updated
keep in mind you DONT need all the audio drivers for each card installed, as they will just eat up resources :P

replace all of your risers

do you have any missing device drivers in device manager?

Here's the thing. I plugged GPU 1 and 2 in and it crashes. Then I try GPU 1 by itself and it works. GPU 2 by itself and it works. So it cant be the risers can it? I didn't change the risers, I just unplugged the other ones from the motherboard. Whichever one I leave plugged in will mine just fine.

haha damn thing, rigs are so finicky sometimes :/   maybe theres a PCI-E port acting up on the MB ? switch around the slots you plug the cards risers into, see what happens :P make sure no dust in the slots also
just keep trying to isolate it,, maybe you can get it to 5 cards, and go from there
and update your system drivers! it could be windows just confused by all the cards :P
also: if you get the time, you can try running a linux OS, then you will know if it works fine on there its just windows being a bitch :P   SMoS, or Hive, NvOS even

OK Update. I think we need to back up. Turns out... even a single card doesn't work. It mines happily for about 5 minutes and then crashes. As opposed to the normal crashing immediately.

So we think the risers are the culprit?

probably :(   you may need to see what exact crash it is too, check windows event log, or watch the bottom of the bluescreen for the error
it will say like "watchdog RPC" or "threads" or something that may help
damn thing, ya maybe SMoS is the solution, you would be surprised how insanely stable it is , tho it does take a bit of setting up off the bat

that would be my guess right now. what risers are you using?

i had an issue with risers when i built my first rig. it wasnt actually the risers, but the USB cables were bad and caused all kinds of issues. fans acting erratically, blue screens, driver crashes, everything.

test one card at a time, one riser at a time. see if you have any good risers/cables. replace any you confirm are bad. maybe even manipulate and bend the usb cable around with the system running to see if it has a repeatable impact on the system behavior

Few weeks back Asus GTX1070 also have similar issue. Temperature error + BSOD after running for a while.

Found out that the SATA cable melted slightly and became faulty. Replaced with a high quality cable and so far no issue. (Uninstalled and reinstalled Driver Version 388.13).

Note: From my experience, Asus cards tends to draw more power that other brands.

Sata cable? As in the sata power cable? I thought we aren't supposed to use those to power risers...

2 gpu per sata  max because gpu mainly draws power from pcie power plug

Can you check your minidump for reason/caused by name? Also bugcheck will be helpful.
Can you tell exact model name of your rosewill psu?
Which bios are you using MIN.F8a or regular one?

Also you can save lot of time, if you just image your windows install and reinstall it via re-imaging (10 min max on ssd over 100Mb network/external storage).

edit
Yes, it does, you just need to flash mining bios. Just checked with amibcp, and regular one doesn have 4g decoding, while mining one have, so I'm pretty sure you troubles may end there, but still provide above info, but I'm still betting on psu (left is regular, right is mining f8a):
https://image.ibb.co/dgzmcH/image.png

can be possible:

1. You can try to change the scheme to powering to/from USB Riser (I assume Riser USB is fine).

2. if you have another PSU with 1000W ++, you can try it.

I had this issue in the past and it was faulty psu, any load on system ( lunching miner) caused gpu fan 100% --> system freeze or restarts

Gpu is dead ,buy new card.

Welp everyone, after a lot of trial and error, I have my first successful launch. Looks like one of the GPU Risers may have been bad. I swear I tested all of these... but I guess one of them died pretty quick. Anyways, I'm gonna let it run while I go to sleep and will check in the morning. Hopefully this fixes everything.

Lesson learned. It's always the damn risers. Even when you swear it isn't.

Allways use excluding technology of diagnostic.
run with one gpu without riser, and check by tgis all cards.
then run with riser one gpu and test by this all risers.
after add one by one gpus and test how works rig

yeah, a bit problem (USB Riser) can be frustrating. actually, about Mining Rig has its complexity.

do not let the problem with USB Riser get into your dreams.

To fix the gpu riser problems the latest versions, the 009s which has all the possible connections to the power supply it is working very well in the mining computers of my colleagues.

They have about 20 mining computers and so far it has been more than one month that every computer is running perfectly without any failure whatsoever until now.

ahh nice man, hope it holds up :)
ya these current market risers are pieces of shit, its crazy how big of an industry this is and we can't even get "brand name" risers yet :(
sometimes they will work off the bat, then catch a piece of dust or static in a few hours and short out and you wont even know till this happens :P

yes, all ways have been tried. actually, the problem is in a riser  ;D. if you see on youtube, the manufacturer produces a very bad riser. I often replace the riser because one or two GPUs are crash and in less than 2 months. I have been tried the Riser version (Ver006, Ver006C, Ver008, Ver008C)

Update. New risers game in... same thing. So either some of the new risers are bad as well or that's not the problem. It started giving the the "claymoremienr... has been blocked from accessing graphics hardware" and then it crashes, goes to the internal video as the main display, and 3 of my 6 graphics cards go unrecognized.

Over 2 weeks now and still not working :/

I guess I'm going to try my 3rd motherboard now (biostar) and see if it's a gigabyte thing.

i would recomend buying an ASUS prime z270-p or the prime h270 plus/csm

i have both and they work perfect.

Did you ever try mining bios on your GA-z270p-d3 with above 4g decoding?