Bitcoin Forum
December 10, 2016, 01:05:17 AM *
News: To be able to use the next phase of the beta forum software, please ensure that your email address is correct/functional.
 
   Home   Help Search Donate Login Register  
Pages: [1]
  Print  
Author Topic: Optimal choice for improving gpu's life cycle without losing mhashes  (Read 1620 times)
BCMan
Hero Member
*****
Offline Offline

Activity: 527



View Profile
January 19, 2012, 10:26:00 PM
 #1

 Whats more optimal - higher fan speed and lesser gpu temp, or auto fan speed with higher (2-3C more - 77C) temp? Can fan fail earlier than gpu with higher speeds?

Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction. Advertise here.
1481331917
Hero Member
*
Offline Offline

Posts: 1481331917

View Profile Personal Message (Offline)

Ignore
1481331917
Reply with quote  #2

1481331917
Report to moderator
1481331917
Hero Member
*
Offline Offline

Posts: 1481331917

View Profile Personal Message (Offline)

Ignore
1481331917
Reply with quote  #2

1481331917
Report to moderator
1481331917
Hero Member
*
Offline Offline

Posts: 1481331917

View Profile Personal Message (Offline)

Ignore
1481331917
Reply with quote  #2

1481331917
Report to moderator
SaintFlow
Sr. Member
****
Offline Offline

Activity: 415


for the love of water


View Profile
January 19, 2012, 10:28:30 PM
 #2

Whats more optimal - higher fan speed and lesser gpu temp, or auto fan speed with higher (2-3C more - 77C) temp?

cooler room and more space between the cards?

which cards?

have you heared of undervolting?

account resurrected
the lurking ends
BCMan
Hero Member
*****
Offline Offline

Activity: 527



View Profile
January 19, 2012, 10:41:06 PM
 #3

Whats more optimal - higher fan speed and lesser gpu temp, or auto fan speed with higher (2-3C more - 77C) temp?

cooler room and more space between the cards?

which cards?

have you heared of undervolting?
Radeon 5770 atm, soon 6870. GPU can be stable @ 960/300 with undervolting?

jake262144
Full Member
***
Offline Offline

Activity: 210


View Profile
January 20, 2012, 06:34:09 PM
 #4

How can anyone tell you whether your particular card has a good enough chip to do that?
You're going to need to test that.

Try dropping the voltage by 0.020 V and running the rig for a day. Repeat until you've found maximal stable undervolt.
Now, there are two choices: you can drop the clock slightly and keep undervolting or you can keep the stable voltage.
If you're actually paying for your energy, dropping the core clock and a serious underclock job are usually the way to go.
Also, undervolting significantly drops core temps.

BCMan
Hero Member
*****
Offline Offline

Activity: 527



View Profile
January 20, 2012, 09:47:52 PM
 #5

 I've tried 1.105 (1.125 is default) and it even doesn't work with 750(850 is stock)/300. GPU stops responding.

Red Emerald
Hero Member
*****
Offline Offline

Activity: 742



View Profile WWW
January 20, 2012, 09:59:28 PM
 #6

I've tried 1.105 (1.125 is default) and it even doesn't work with 750(850 is stock)/300. GPU stops responding.
When you are tweaking for undervolting, your goal is higher MH/s/wattJ, not highest MH/s.  Get a watt meter if you are serious about this.

jake262144
Full Member
***
Offline Offline

Activity: 210


View Profile
January 20, 2012, 10:41:33 PM
 #7

I've tried 1.105 (1.125 is default) and it even doesn't work with 750(850 is stock)/300. GPU stops responding.

Sorry to hear this :<
It's a total crapshoot, some GPUs have just higher quality chips than others.
BCMan
Hero Member
*****
Offline Offline

Activity: 527



View Profile
January 20, 2012, 10:54:02 PM
 #8

 Actually I did it in wrong way. I oced it in AMD overdrive trough ATI Tray Tools (stupid ATT overclocking tool cant up freq more than 750 mhz and lower than 500) and somehow overdrive conflicted with different frequency, even with 1.124v. I've found a workaround. Now I'm running 960/300/1.105v and its working! Alsogpu is a bit colder now, not that much tho, lesser on 1-1.5C. Wondering how far I can go with 960/300.

jake262144
Full Member
***
Offline Offline

Activity: 210


View Profile
January 20, 2012, 11:15:26 PM
 #9

Way to go BCMan!
Give that card a day. If it's still perfectly stable tomorrow, undervolt again.
If you manage to repeat this 3 or 4 times the card should be about 5 °C colder taking care of the thermal issue you raised earlier.
RandyFolds
Sr. Member
****
Offline Offline

Activity: 434



View Profile
January 21, 2012, 12:41:03 AM
 #10

Actually I did it in wrong way. I oced it in AMD overdrive trough ATI Tray Tools (stupid ATT overclocking tool cant up freq more than 750 mhz and lower than 500) and somehow overdrive conflicted with different frequency, even with 1.124v. I've found a workaround. Now I'm running 960/300/1.105v and its working! Alsogpu is a bit colder now, not that much tho, lesser on 1-1.5C. Wondering how far I can go with 960/300.

There are some weird resonance things that happen...maybe someone more technically adept than I can explain them. Point being that if you are tweaking clock settings, hitting an unstable setting doesn't mean you can't pass it and find a sweet spot. Just making up numbers here for an example; shit may hit the fan at 950mhz core and be fine at 940 and 960mhz.

▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓
▓▓ ONEDICE.ME ▓▓▓▓▓ BEST DICE EXPERIENCE ▓▓▓▓ PLAY OR INVEST ▓▓▓▓▓▓
▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓
BCMan
Hero Member
*****
Offline Offline

Activity: 527



View Profile
January 21, 2012, 10:44:21 AM
 #11

 I'm testing it almost a day already with 960/300/1.005v. Shocked Rock solid stability. Incredible how much it can be undervolted, even with quite high OCing! Funny thing that with 1.000v it doesn't work, but with 1.005 working without any issues. And temp is at 71C, comparing to 77 with default voltage. Thats just great.

P4man
Hero Member
*****
Offline Offline

Activity: 504



View Profile
January 21, 2012, 10:58:52 AM
 #12

To get back to your original question; running at high fan speeds is more likely to kill the fan prematurely. Running at low fan speeds (assuming that results in high temperatures) is more likely to cause premature failure of the GPU. Its impossible to put numbers on it, premature could mean 15 years instead of 150, it could mean 2 months instead of 20. But replacing a fan is fairly cheap and easy, replacing the GPU is less so.

Settle for a compromise that you feel makes sense; to me that means keeping the GPU below 70C, whatever fan speed that requires. But Im not going to double my fan speed to max only to lower temps from 65C to 55C.

jake262144
Full Member
***
Offline Offline

Activity: 210


View Profile
January 21, 2012, 12:12:25 PM
 #13

Right on, P4.  I'm glad you raised the often overlooked point of diminishing returns at high fan speeds.
Just keep in mind that "below 70°C" is an extreme oversimplification:
   The 69xx series can never be expected to run that cold. They are all flawed products, having been meant to be manufactured at 32nm instead of 40, and will run more like high 70s - low 80s.

@BCMan, now you're talking dude Smiley Welcome to the land of safe and efficient mining.
BCMan
Hero Member
*****
Offline Offline

Activity: 527



View Profile
January 21, 2012, 11:06:16 PM
 #14

To get back to your original question; running at high fan speeds is more likely to kill the fan prematurely. Running at low fan speeds (assuming that results in high temperatures) is more likely to cause premature failure of the GPU. Its impossible to put numbers on it, premature could mean 15 years instead of 150, it could mean 2 months instead of 20. But replacing a fan is fairly cheap and easy, replacing the GPU is less so.
I agree, but running @ 100% fan speed I'm getting only 2-3C lesser temp. Not very effective I would say and my room becomes a server room. Maybe add some decent external fan, like this one:
 

jjshabadoo
Hero Member
*****
Offline Offline

Activity: 519


View Profile
January 21, 2012, 11:29:09 PM
 #15

hahahaha, good thread and that post was funny as hell.

Good luck with the under volting, I wish we would do more sticky's/guides on this site with info about this stuff, that wiki article for mining hardware is so out of date it's not even valuable.

At least give people an idea of what they can shoot for. If you've ever oc'd or changed anything from stock, you know even though it's the same product, there are always different batches/chips, etc. but it still helps because there is usually a fairly accurate average.
Pages: [1]
  Print  
 
Jump to:  

Sponsored by , a Bitcoin-accepting VPN.
Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!