[Spongeguard]

With 0 errors

What are the chances that my new HD 5770 is faulty/defective?

 

[wumpus]

almost 0

that also tells you that it works properly as well.

 

[Lune]

95C 6 hours... sounds like you are havin a barbeque.. dont do that again:>

 

[Scorpion87]

Quote:

Originally Posted by Lune 95C 6 hours... sounds like you are havin a barbeque.. dont do that again:>

This!

 

[mth91]

Boil me up a cup of tea there chap.

 

[Volvo]

That's bad.

Do you have an overclock?
If no, sounds like you've got a defective cooler/fan speed isn't ramping.

I once left my fan at the minimum speed on manual mode, played TF2 and my 5770 got so hot it throttled.

 

[stormcrow]

it most likely lost half of its life in the process as well.

congratulations!

 

[Wishmaker]

Quote:

Originally Posted by stormcrow it most likely lost half of its life in the process as well. congratulations!

+1. Why would you do that? Run 6hrs at 95 Celsius??

 

[whitingnick]

Not a great thing to do to you card.

 

[ENTERPRISE]

Lol so many people get scared when a GPU hits 90C...That is nothing. Sure the cooler the card the better but I think so many people forget that GPU's by thermal design are able to take 100C+ My fermi at 100% GPU usage in folding gets to 92C...that is nothing. Card does not break a sweat and keeps on chugging.

I think people are way to sensitive with GPU's.

 

[Wishmaker]

Quote:

Originally Posted by ENTERPRISE Lol so many people get scared when a GPU hits 90C...That is nothing. Sure the cooler the card the better but I think so many people forget that GPU's by thermal design are able to take 100C+ My fermi at 100% GPU usage in folding gets to 92C...that is nothing. Card does not break a sweat and keeps on chugging. I think people are way to sensitive with GPU's.

I disagree. People are sensitive to HOT COMPONENTS. Heat was, is and always will be the enemy for computer parts

. Sure, TJMAX on my I7 is 100 celsius, does this mean that I can keep the thing at 95 for 12 hours without reducing its life? Same with a GPU, the hotter it is, the probability of reducing its life or having it to fail is greater.

 

[ENTERPRISE]

Perhaps I am just more hardcore but the heat from the GPU IF it is actually decreasing the lifespan...will not make any difference. The card will still last a good amount of time. I will have upgraded my GPU way before the 470 dies out due to heat exposure. The only thing I get worried about heat wise is the CPU and any other confined components but as the GPU is designed to take a beating with regards to heat it is something I worry about the least.

 

[Lune]

We got it, you don't care about your components :>

 

[Setzer]

I've seen my card run at 105C (!!!!), before I cleaned all the dust out of it, it used to be in the early 90's when playing demanding games, and Crysis would make it crash and BSOD sometimes.
I cleaned out the dust, and now it maxes out at 79C AT MOST, only FurMark can get it to 86C.

This is on 50% fan speed.

And heck, my GPU has been running for 2 years now.

 

[angrysasquatch]

Quote:

Originally Posted by ENTERPRISE Lol so many people get scared when a GPU hits 90C...That is nothing. Sure the cooler the card the better but I think so many people forget that GPU's by thermal design are able to take 100C+ My fermi at 100% GPU usage in folding gets to 92C...that is nothing. Card does not break a sweat and keeps on chugging. I think people are way to sensitive with GPU's.

Yup, I overclocked my old 8600gt so it would run at 110* load (trying to kill it). It really just didn't give a damn, no artifacts, no instability, and still lives.

Quote:

Originally Posted by Wishmaker I disagree. People are sensitive to HOT COMPONENTS. Heat was, is and always will be the enemy for computer parts . Sure, TJMAX on my I7 is 100 celsius, does this mean that I can keep the thing at 95 for 12 hours without reducing its life? Same with a GPU, the hotter it is, the probability of reducing its life or having it to fail is greater.

You're making an assumption. CPUs do sometimes have to have the overclock backed off or voltage upped for stability in relation to what it used to be able to clock at because of electromigration. I have yet to see a case in which a GPU does the same.

 

[Spongeguard]

Please note that the card hit 95c because I was running shader complexity 8.
When I run furmark ni xtreme burn testing mode it hits 82c, and when gaming it hits a max of 71c.

And no, I did not take half of the card's lifetime in the process. Cards are designed to take heat.

But seriously, instead of criticizing me, can you just answer my initial question.
What are the chances of a faulty card making it through 6 hours of OCCT GPU at 95c?

 

[Peroxide]

None of you have laptops or never bothered to install GPUz on them?

90+ is expected with a lot of gaming laptops..

 

[stormcrow]

Quote:

Originally Posted by Spongeguard But seriously, instead of criticizing me, can you just answer my initial question. What are the chances of a faulty card making it through 6 hours of OCCT GPU at 95c?

sorry i was just messing with ya. didn't think you'd take it seriously.

to answer your question... yes, i would say you are pretty stable if you managed to do that. although i would rather put my trust on furmark rather than occt for stability testing video cards.

in any case, 6 hours is a LONG time to be putting a gpu at 95C. that's 6 straight hours of constant 3D mode and having your video card above 90C during those straight 6 hours could potentially--i said potentially--damage something, somewhere in the card. maybe not the GPU itself, then perhaps the RAM or the VRMs. i imagine the VRMs are way above 100C during the first few minutes. you ran it for 6 hours straight.

it's not the same as playing a game for 6 hours straight. because when you play a game, you switch back and forth into 2D & 3D mode (when you pause, when you access menus). thus giving the GPU some time to rest. temps also fluctuate during gaming from lows to highs, vice versa.

 

[Lune]

Quote:

Originally Posted by Spongeguard Please note that the card hit 95c because I was running shader complexity 8. When I run furmark ni xtreme burn testing mode it hits 82c, and when gaming it hits a max of 71c. And no, I did not take half of the card's lifetime in the process. Cards are designed to take heat. But seriously, instead of criticizing me, can you just answer my initial question. What are the chances of a faulty card making it through 6 hours of OCCT GPU at 95c?

That is like "I cba cooling my card" to me.

and aint it kinda obvious? ofc your card is not FAULTY if you didnt crash for 6 hours, sigh.

 

[Spongeguard]

Quote:

Originally Posted by stormcrow sorry i was just messing with ya. didn't think you'd take it seriously. to answer your question... yes, i would say you are pretty stable if you managed to do that. although i would rather put my trust on furmark rather than occt for stability testing video cards. in any case, 6 hours is a LONG time to be putting a gpu at 95C. that's 6 straight hours of constant 3D mode and having your video card above 90C during those straight 6 hours could potentially--i said potentially--damage something, somewhere in the card. maybe not the GPU itself, then perhaps the RAM or the VRMs. i imagine the VRMs are way above 100C during the first few minutes. you ran it for 6 hours straight. it's not the same as playing a game for 6 hours straight. because when you play a game, you switch back and forth into 2D & 3D mode (when you pause, when you access menus). thus giving the GPU some time to rest. temps also fluctuate during gaming from lows to highs, vice versa.

Well, thanks for the explanation.
I understand that I may have reduced the life, but don't GPU's have life expectancy of 2-3 years? I'll have a new gpu LONG before that.

Quote:

Originally Posted by Lune That is like "I cba cooling my card" to me. and aint it kinda obvious? ofc your card is not FAULTY if you didnt crash for 6 hours, sigh.

And yes, I suppose it IS obvious. I'm just experiencing so many graphics issues that I did not experience with my nVidia cards. I suppose it's natural for ATi cards.