[th3illusiveman]

Hey guys, does having a high ASIC quality affect VRM temperature? I ask because my VRMs are getting very hot with low voltage. (Well, low-ish voltage) At 1218mv i see Vrm temps in the high 80's and low 90's yet when i bump it down to 1174mv the temps rarely break 74c. i know pumping 44mv more isn't a low amount of voltage but a 15c difference between the two is pretty steep. My ASIC quality is 87%

During all this the GPU core does not exceed 54c.

I read somewhere that people with ASIC quality below 60 have hotter cores with cooler running VRMs at higher voltages where as people with high ASIC quality have cooler cores with hotter VRMs at lower voltage. All his was from afew forum users so i don't know how solid their posts are.

 

[thehammer007]

Higher ASIC quality generally means you need less voltage to get to a certain amount of clock speed. So in theory, lower ASIC = higher VRM temps.

Some cards like the Powercolor Vortex 2 does not have proper VRM cooling, same thing with the XFX Double D. I don't think your VRMs are too hot, if you feel that it is too hot, get an Antec Spotcool or a backplate.

 

[glina]

Quote:

Originally Posted by thehammer007

Higher ASIC quality generally means you need less voltage to get to a certain amount of clock speed. So in theory, lower ASIC = higher VRM temps.

Some cards like the Powercolor Vortex 2 does not have proper VRM cooling, same thing with the XFX Double D. I don't think your VRMs are too hot, if you feel that it is too hot, get an Antec Spotcool or a backplate.

It's the opposite on AMD cards according to this:
Quote:

Update (2012.01.23): the ASIC quality detection is probably based on the GPU voltage. Here what AMD's Dave Baumann says:

Actually, it does the opposite! We scale the voltage based on leakage, so the higher leakage parts use lower voltage and the lower leakage parts use a higher voltage - what this is does narrow the entire TDP range of the product.

Everything is qualified at worst case anyway; all the TDP calcs and the fan settings are completed on the wors case for the product range.

source: http://www.geeks3d.com/20120122/test-asic-quality-of-geforce-gpus/

This would mean that in fact higher ASIC chips are the leaky ones, and the lower ASIC ones are less power hungry.
This certainly seems to be the case with mine. I have 87% asic and the card is extremely power hungry, which could also explain the hot VRMs.

As what this means for overclocking: I see people reaching 1250-1300MHz with any asic value between 60 and 90% so there certainly more to that than a simple "ASIC quality" rating.

 

[tsm106]

Quote:

Originally Posted by glina

Quote:

Originally Posted by thehammer007

Higher ASIC quality generally means you need less voltage to get to a certain amount of clock speed. So in theory, lower ASIC = higher VRM temps.

Some cards like the Powercolor Vortex 2 does not have proper VRM cooling, same thing with the XFX Double D. I don't think your VRMs are too hot, if you feel that it is too hot, get an Antec Spotcool or a backplate.

It's the opposite on AMD cards according to this:
Quote:

Update (2012.01.23): the ASIC quality detection is probably based on the GPU voltage. Here what AMD's Dave Baumann says:

Actually, it does the opposite! We scale the voltage based on leakage, so the higher leakage parts use lower voltage and the lower leakage parts use a higher voltage - what this is does narrow the entire TDP range of the product.

Everything is qualified at worst case anyway; all the TDP calcs and the fan settings are completed on the wors case for the product range.

source: http://www.geeks3d.com/20120122/test-asic-quality-of-geforce-gpus/

This would mean that in fact higher ASIC chips are the leaky ones, and the lower ASIC ones are less power hungry.
This certainly seems to be the case with mine. I have 87% asic and the card is extremely power hungry, which could also explain the hot VRMs.

As what this means for overclocking: I see people reaching 1250-1300MHz with any asic value between 60 and 90% so there certainly more to that than a simple "ASIC quality" rating.

Don't use Dave's explanation he's actually wrong. Go look up the definition of cpu leakage and you'll see everything Dave said is the opposite. You can even see the truth in the way tahiti asic works. High asic cards load with very low voltage. Low asic cards load with higher voltage comparatively. Then when you examine the droop, low asic droop like crazy and high asic do not. So please stop posting the Dave's rubbish.

 

[glina]

Quote:

Originally Posted by tsm106

Don't use Dave's explanation he's actually wrong. Go look up the definition of cpu leakage and you'll see everything Dave said is the opposite. You can even see the truth in the way tahiti asic works. High asic cards load with very low voltage. Low asic cards load with higher voltage comparatively. Then when you examine the droop, low asic droop like crazy and high asic do not. So please stop posting the Dave's rubbish.

Lets find out who's telling the truth then.

What are your ASIC values and droop?
Mine's 87% and droop 0.09V in Furmark. (@1100mhz; 1.087V)

 

[tsm106]

Are you feeling daft? What the heck do you think it means to leak voltage? Go Google it for yourself and learn something.

 

[Jopel]

ASIC quality of 64,9%.
My 7970 DC2T does 1260/6800(1700MHz) @ 1300Mv

 

[th3illusiveman]

So far i've been game stable at 1225Mhz with 1175mV, hoping it holds up until the end of the week.

 

[Derpinheimer]

ASIC 88% - 7950
Runs 1200 @ 1.137v, vDroop to 1.030-1.050 depending on load

I wonder if this theory is correct, as even having a Vapor-X card, which are notorious for bad VRM temps, mine is still awful with a card that can run 1200 at a pretty incredible voltage.

 

[sugarhell]

84,2 % ASIC quality. 1.07 stock voltage

 

[th3illusiveman]

Quote:

Originally Posted by Derpinheimer

ASIC 88% - 7950
Runs 1200 @ 1.137v, vDroop to 1.030-1.050 depending on load

I wonder if this theory is correct, as even having a Vapor-X card, which are notorious for bad VRM temps, mine is still awful with a card that can run 1200 at a pretty incredible voltage.

Well they say they are rated for 120c and some users have been unwise enough to let their go to 140c for afew minutes and still have a live card. So maybe that's why some GPU manufacturers don't care too much about cooling them well (as well as aftermarket manufacturers as i've come to learn the hard way

) The MSI Lightning has excellent VRM cooling but the fans are noisy.

So i think they are pretty tough and can handle alot of heat, still though "i know that feel bro"

i hate high temps and seeing anything running at 90c rustles my jimmies.

 

[glina]

Quote:

Originally Posted by tsm106

Are you feeling daft? What the heck do you think it means to leak voltage? Go Google it for yourself and learn something.

Perhaps you have trouble understanding what I quoted. GPU-Z has no way of measuring the CURRENT leakage, and so gives its rating based only on the stock voltage readout.
What Dave Baumann said, is that AMD is supposedly binning their chips the other way around in order to have the same TDP on all of them, even if it means for some to have an unnecesarily high voltage. It's pretty clear that chips with higher leakage require higher voltages, but this is not what is discussed here.

 

[tsm106]

Quote:

Originally Posted by glina

Quote:

Originally Posted by tsm106

Are you feeling daft? What the heck do you think it means to leak voltage? Go Google it for yourself and learn something.

Perhaps you have trouble understanding what I quoted. GPU-Z has no way of measuring the CURRENT leakage, and so gives its rating based only on the stock voltage readout.
What Dave Baumann said, is that AMD is supposedly binning their chips the other way around in order to have the same TDP on all of them, even if it means for some to have an unnecesarily high voltage. It's pretty clear that chips with higher leakage require higher voltages, but this is not what is discussed here.

Quote:

higher leakage parts use lower voltage and the lower leakage parts use a higher voltage

 

[glina]

Yes. One is physics, the other one is how AMD bins gpus (supposedly). Think about it - there's only one way to have equal TDP on two chips with different leakages, and it is to run the lower leakage one with higher volts, even if it doesn't require this.

 

[tsm106]

Quote:

Originally Posted by glina

Yes. One is physics, the other one is how AMD bins gpus (supposedly). Think about it - there's only one way to have equal TDP on two chips with different leakages, and it is to run the lower leakage one with higher volts, even if it doesn't require this.

That's not what's happening lol.

 

[glina]

Quote:

Originally Posted by tsm106

That's not what's happening lol.

Perhaps in your world, where amd drivers throttle furmark

 

[tsm106]

Quote:

Originally Posted by glina

Quote:

Originally Posted by tsm106

That's not what's happening lol.

Perhaps in your world, where amd drivers throttle furmark

So you got nothing and resort to ad hominem?

You contradicted Bauman. What is this separation from Baumans coked world with the real world of physics?? Everything known about leakage is opposite to what he wrote. LMAO. Waste of time.

 

[glina]

Quote:

Originally Posted by tsm106

So you got nothing and resort to ad hominem?

You contradicted Bauman. What is this separation from Baumans coked world with the real world of physics?? Everything known about leakage is opposite to what he wrote. LMAO. Waste of time.

Thats what you're failing to understand. Baumann knows leakage. It's just the binning they do the opposite way.
BTW, most GHZ edition 7970 have asic quality of 60-70% - this is because these are the least power hungry chips (least leaky).
But believe whatever you want.

 

[sugarhell]

Lol higher ASIC use lower voltage. One of my card is 88% with 1.07 stock voltage. The others is 70 % with the common 1.17 voltage.

Its pretty easy to understand higher asic lower voltage. If you think otherwise then just go play tetris with an oc 7970

 

[tsm106]

Quote:

Originally Posted by glina

Quote:

Originally Posted by tsm106

So you got nothing and resort to ad hominem?

You contradicted Bauman. What is this separation from Baumans coked world with the real world of physics?? Everything known about leakage is opposite to what he wrote. LMAO. Waste of time.

Thats what you're failing to understand. Baumann knows leakage. It's just the binning they do the opposite way.
BTW, most GHZ edition 7970 have asic quality of 60-70% - this is because these are the least power hungry chips (least leaky).
But believe whatever you want.

I'm feeling ******ed now after this discourse.