Bitcoin Forum
May 11, 2024, 08:43:19 AM *
News: Latest Bitcoin Core release: 27.0 [Torrent]
 
   Home   Help Search Login Register More  
Pages: [1]
  Print  
Author Topic: Ethereum strange GPU problem when unpluging monitor  (Read 52 times)
Gaweu (OP)
Newbie
*
Offline Offline

Activity: 2
Merit: 0


View Profile
January 18, 2018, 11:00:55 AM
 #1

English is not my native so sorry if this will hurt Your eyes. Smiley

So I have 6 GPU rig.
mobo gigabyte h110 d3a
6x sapphire rx 580 nitro+ 8gb hynix
Im using claymore for eth single mining so far.

When I built this rig and i was setting it up everything was fine. There was almost no problem at all beside that windows was running very slow ( on both 4gb and 8gb ram setup ). By slow i mean it didnt sometimes respond for things like clicking on start button with LMB or going to windows 10 settings. But when i was for example clicking RMB on windows start button then it worked just fine. There was also problem with restarting sometimes becuase when i tryed to do that i was getting timeout of Task Host manager if im not wrong or something like that. ( I looked for all the solutions on google nothing helped me ). I reinstalled windows Twice and that didnt help either. I was looking for some DAG fixes ( because there is the problem most of the time ) that didnt help either.

So my main problem is DAG buffer or something with basicly buffor. When i launch Claymore it just freeze  on creating buffor. Sometimes it will pass that but wont create DAG for GPU0 and it will start mining but after like 1 hashrate that is correct for the card it will drop down to 0mh/s and stay that way untill claymore will reset because of error.

I made adjustments and i swap risers and even cards and i noticed strange thing ( to bad it was like after 2-3 days of me getting pissed AF! But later about that )
Problem was always with card that is in very left side just above 2 PSU, 1200 eVGA plat and some 650 gold that i added recently ( problem was there before this additional one too ) because i was thinking that maybe its because there is not enough power even when my thing to test watts showed that im using 1000w from the wall.

So when I was taking rig from place when i keep them ( little room with AC where You can check them only with remote desktop  ) to place where i can plug monitor and rest. I was getting crazy with setting all up like 10 times per day because every thing was GREAT! no problems at all. So i was like ok ill hold it for like hour or so to just be sure.  Then when all was prefect i was like ok its time to move it back. As soon as i putted rig in that room with AC i didnt work anymore so every time i was like switching risers move cards a bit etc etc... I mean it worked sometimes but it was like 2 times out of 10

To make story short after like 20-25 hours total last few days i was getting NUTS and tripping that i overclocked card to much etc! And looking for help on google etc i noticed that when ever I unplug monitor form RIG that one card just fail. It cant create DAG, or what ever and its minnning with 0mh/s. When ever i plug monitor back all is great. Even when software is running and its stuck on creating DAG when i plug monitor its just go like nothing was wrong sec ago. There is no difference what card is on that spot. I pluged XFX rx580 and it was same deal. The other thing that i noticed but i might be wrong on this one is that i think my bios was like broken or something after some time becuase i could not load in to it either with spaming del ( fast boot was disabled ) or by shift + restart windows and using that option to go in to bios on reboot. Maybe that will help someone with identyfication of my problem.

Any idea what is that problem? Is it about some bad electricity flow or how to say that? Im using aluminium frame for rig. But if that would be case why only 1 gpu fail?

EDIT:
I was reading also about some dummy plugs for GPU's but i guess its not my problem becuase my rig can run fine with out those. But that one GPU just fails 80% of time.
1715416999
Hero Member
*
Offline Offline

Posts: 1715416999

View Profile Personal Message (Offline)

Ignore
1715416999
Reply with quote  #2

1715416999
Report to moderator
The grue lurks in the darkest places of the earth. Its favorite diet is adventurers, but its insatiable appetite is tempered by its fear of light. No grue has ever been seen by the light of day, and few have survived its fearsome jaws to tell the tale.
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction.
1715416999
Hero Member
*
Offline Offline

Posts: 1715416999

View Profile Personal Message (Offline)

Ignore
1715416999
Reply with quote  #2

1715416999
Report to moderator
1715416999
Hero Member
*
Offline Offline

Posts: 1715416999

View Profile Personal Message (Offline)

Ignore
1715416999
Reply with quote  #2

1715416999
Report to moderator
1715416999
Hero Member
*
Offline Offline

Posts: 1715416999

View Profile Personal Message (Offline)

Ignore
1715416999
Reply with quote  #2

1715416999
Report to moderator
dschwartz0815
Newbie
*
Offline Offline

Activity: 29
Merit: 0


View Profile
January 18, 2018, 01:23:01 PM
 #2

Have you tried making sure drivers are up to date?

Your GPU drivers should be here: https://www.sapphiretech.com/download.asp?lang=eng

Also check for motherboard BIOS updates and drivers: http://www.gigabyte.us/Motherboard/GA-H110-D3A-rev-10#support-dl
Souran
Jr. Member
*
Offline Offline

Activity: 59
Merit: 4


View Profile
January 18, 2018, 01:31:52 PM
 #3

I had a similar issue, at first I thought it was because Im running 12 gpu per rig and its just drivers arent stable.

But eventually found out that I needed a dummy plug or a monitor to be attached to the miner.
It didnt make sense, because I had other rigs running fine.
Im using something like these now:
https://www.aliexpress.com/item/Built-in-1080p-Chipset-Converter-HDMI-to-VGA-with-Audio-Cable-Micro-Mini-HDMI-Male-Adapter/1785843152.html

They are cheaper than the dummy plugs and works just as good.

Other issue I had, was undervolting, one of my cards was acting funny, all my other identical cards were running with the same undervolting (100+ cards), but this one needed an extra 50mv to be stable even at lower clocks.
Kapz786
Hero Member
*****
Offline Offline

Activity: 697
Merit: 503



View Profile
January 18, 2018, 01:40:02 PM
 #4

I had a similar issue, at first I thought it was because Im running 12 gpu per rig and its just drivers arent stable.

But eventually found out that I needed a dummy plug or a monitor to be attached to the miner.
It didnt make sense, because I had other rigs running fine.
Im using something like these now:
https://www.aliexpress.com/item/Built-in-1080p-Chipset-Converter-HDMI-to-VGA-with-Audio-Cable-Micro-Mini-HDMI-Male-Adapter/1785843152.html

They are cheaper than the dummy plugs and works just as good.

Other issue I had, was undervolting, one of my cards was acting funny, all my other identical cards were running with the same undervolting (100+ cards), but this one needed an extra 50mv to be stable even at lower clocks.

Spot on - most of the instability issue can be solved with slightly more voltage as well dummy plugs. If you have a read over the XMR Cast thread a lot of issues are resolved by this.
Gaweu (OP)
Newbie
*
Offline Offline

Activity: 2
Merit: 0


View Profile
January 18, 2018, 07:15:39 PM
 #5

Thanks for answers. Yea i tryed all the drivers and such.

I tought that dummy plug gonna be easiest solution at the end.

Funny thing is that its not about GPU itself but the place on the rig. If ill put other GPU there no matter of company or even which PCIe it is connected to it will just fail there Smiley

Ok ill try that dummy plug.

Thanks
Pages: [1]
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!