Bitcoin Forum
April 27, 2024, 05:23:34 AM *
News: Latest Bitcoin Core release: 27.0 [Torrent]
 
   Home   Help Search Login Register More  
Warning: One or more bitcointalk.org users have reported that they strongly believe that the creator of this topic is a scammer. (Login to see the detailed trust ratings.) While the bitcointalk.org administration does not verify such claims, you should proceed with extreme caution.
Pages: [1]
  Print  
Author Topic: Help with proper clocks.  (Read 1049 times)
5stringer (OP)
Member
**
Offline Offline

Activity: 185
Merit: 10



View Profile
July 28, 2013, 05:00:54 AM
 #1

I'm using cgminer 3.3.1, on windows 7 64-bit. I have 4 6870s ( 2 awaiting risers and not in use. Yet. )   Now when I run MSI afterburner, GPU1 useage % tops at 89% while GPU2 useage tops at 99%. Sometimes it crashes and it reads 200% useage on GPU2.

I'm just curious to know if anyone in the community would know what would cause this?

I'd also like to know what would be the best Core Clock and Memory Clock settings for them. (Mining comparison site kind of throws me off. whether its 6870 x2 or 6870 x4 settings)

Also Curious if any certain Linux distros would be easier to mine with?

As always, Thank /each/ and /every/ One of you who gives me feedback.



1714195414
Hero Member
*
Offline Offline

Posts: 1714195414

View Profile Personal Message (Offline)

Ignore
1714195414
Reply with quote  #2

1714195414
Report to moderator
1714195414
Hero Member
*
Offline Offline

Posts: 1714195414

View Profile Personal Message (Offline)

Ignore
1714195414
Reply with quote  #2

1714195414
Report to moderator
Remember that Bitcoin is still beta software. Don't put all of your money into BTC!
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction.
PEBKAC
Newbie
*
Offline Offline

Activity: 28
Merit: 0



View Profile
July 28, 2013, 03:44:57 PM
 #2

Well, first it would be good to find out what the temperatures of your cards are.
It may just as well be an overheating issue. Have you tried installing Speedfan or some other temperature monitoring tool?

Also, are your GPU's overclocked?
Overclocking increases temperature and increases the amount of power the GPU needs, wich in turn heats up the voltage regulators and could cause instability.

Also, a powersupply that cannot keep up could also create this kind of problem.
5stringer (OP)
Member
**
Offline Offline

Activity: 185
Merit: 10



View Profile
July 28, 2013, 09:26:31 PM
 #3

Well, first it would be good to find out what the temperatures of your cards are.
It may just as well be an overheating issue. Have you tried installing Speedfan or some other temperature monitoring tool?

Also, are your GPU's overclocked?
Overclocking increases temperature and increases the amount of power the GPU needs, wich in turn heats up the voltage regulators and could cause instability.

Also, a powersupply that cannot keep up could also create this kind of problem.

Under load, gpu1 59c ,  gpu2 70c.

No overclocking, all stock speeds. trying to figure out the best setting for the cards to last a while.

under MSI AB, GPU1 usage % 89, GPU2 usage % 99.

I'm not going to lie, I'm a cgminer newb But to me it seems as if gpu2 crashes after so long.

Any help is appreciated
chinesechicken
Newbie
*
Offline Offline

Activity: 33
Merit: 0



View Profile
July 29, 2013, 02:14:30 AM
 #4

all depends on what currency your mining  Huh Huh Huh


for example in BTC;

in AF make sure you have the "sync similar gpu option" UNCHECKED

keep the voltage at stock, push powertune all the way up, drop the vram and far as you can
push the core clock in increments of 5 and test for 12-24 hours. if the gpu is stable continue to push addition 5 and repeat until your GPU crashes then you know your cards clock. you can then start to drop the voltage little by little testing yet again 12-24 hours until your card crashs

OC testing is rigorous and time consuming. usually requires a lot of babysitting your rigs. but once you find the sweet spot your rig is pretty much automated
5stringer (OP)
Member
**
Offline Offline

Activity: 185
Merit: 10



View Profile
July 29, 2013, 03:41:53 AM
 #5

all depends on what currency your mining  Huh Huh Huh


for example in BTC;

in AF make sure you have the "sync similar gpu option" UNCHECKED

keep the voltage at stock, push powertune all the way up, drop the vram and far as you can
push the core clock in increments of 5 and test for 12-24 hours. if the gpu is stable continue to push addition 5 and repeat until your GPU crashes then you know your cards clock. you can then start to drop the voltage little by little testing yet again 12-24 hours until your card crashs

OC testing is rigorous and time consuming. usually requires a lot of babysitting your rigs. but once you find the sweet spot your rig is pretty much automated

I Merge mine on the guild. ( BTC & NMC ) cgminer 3.3.1

"in AF make sure you have the "sync similar gpu option" UNCHECKED" - I take this as a referral to Afterburner? 

"keep the voltage at stock, push powertune all the way up, drop the vram and far as you can
push the core clock in increments of 5 and test for 12-24 hours. if the gpu is stable continue to push addition 5 and repeat until your GPU crashes then you know your cards clock. you can then start to drop the voltage little by little testing yet again 12-24 hours until your card crashs" - Is this cgminer related? quite the cgminer noob Undecided
BitCoiner2012
Sr. Member
****
Offline Offline

Activity: 364
Merit: 250



View Profile
July 29, 2013, 03:50:02 AM
 #6

all depends on what currency your mining  Huh Huh Huh


for example in BTC;

in AF make sure you have the "sync similar gpu option" UNCHECKED

keep the voltage at stock, push powertune all the way up, drop the vram and far as you can
push the core clock in increments of 5 and test for 12-24 hours. if the gpu is stable continue to push addition 5 and repeat until your GPU crashes then you know your cards clock. you can then start to drop the voltage little by little testing yet again 12-24 hours until your card crashs

OC testing is rigorous and time consuming. usually requires a lot of babysitting your rigs. but once you find the sweet spot your rig is pretty much automated

I Merge mine on the guild. ( BTC & NMC ) cgminer 3.3.1

"in AF make sure you have the "sync similar gpu option" UNCHECKED" - I take this as a referral to Afterburner? 

"keep the voltage at stock, push powertune all the way up, drop the vram and far as you can
push the core clock in increments of 5 and test for 12-24 hours. if the gpu is stable continue to push addition 5 and repeat until your GPU crashes then you know your cards clock. you can then start to drop the voltage little by little testing yet again 12-24 hours until your card crashs" - Is this cgminer related? quite the cgminer noob Undecided



Cgminer is great and easy. There is a GPU management button, choose the device, and you can alter clock and mem speeds, as well as fan settings. Alarmingly easy!

BTC Long.
5stringer (OP)
Member
**
Offline Offline

Activity: 185
Merit: 10



View Profile
July 29, 2013, 04:54:03 AM
 #7


"Cgminer is great and easy. There is a GPU management button, choose the device, and you can alter clock and mem speeds, as well as fan settings. Alarmingly easy!"


These are all things I am aware of. Like the OP stated, "GPU1 usage % 88. GPU2 usage % 99" What would cause this with stock Core and Memory clocks? ( Haven't touched anything to do with OC'ing.
5stringer (OP)
Member
**
Offline Offline

Activity: 185
Merit: 10



View Profile
July 29, 2013, 12:35:22 PM
 #8

I should also mention the cards I am using are. 1x HD-687A-ZH HD-687A-ZHFC V1.4 and 1x HD-687A-ZN HD-687A-ZNFC V2.0 ( Also have two more 687a-znfc v1.1 that aren't in use yet. )


Not sure if these 6870s different Version #'s makes any difference or not.
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!