I thinkt the point is the normal user would be happier with a quieter fan. The Enthusiast is going to edit the fan profiles to 100% and probably overclock anyway. Pretty much like the people on the EVGA are doing.
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
EVGA user has done a GTX480 Review
- Thread starter AWPC
- Start date
I thinkt the point is the normal user would be happier with a quieter fan. The Enthusiast is going to edit the fan profiles to 100% and probably overclock anyway. Pretty much like the people on the EVGA are doing.
Soldato
- Joined
- 1 Apr 2010
- Posts
- 3,034
I think Enthusiast arent so phased by the heat and noise as they normally have watercooling or a really good aircooled setup. Enthusiasts are more concerned with performance.
There is a video which I cant seem to find at the moment which shows a 480 with a 92 or 120mm fan pointed at the heatsink. The temps and noise levels are allot lower with the extra fan there. They did a side by side with a 5870 and the 480gtx on stock and with extra fan. With the extra fan it was really cool and quiet. If I can find it I will post it up.
Soldato
- Joined
- 1 Apr 2010
- Posts
- 3,034
great that would be appreciated
I think I might of bookmarked the link on my pc at home.
Associate
I said nothing about it not being an apples to apples comparison. I agree that the 5XXX series is quieter but what I was saying was that running a db meter right next to the fan is not going to offer as relevant a test as running within a closed case in normal operating circumstances with the db meter sat where a user would sit. It gives you the worst case, not a real world case.
That's the kind of test that I'd be interested in, not "look how loud it is with a microphone right next to it!!" I'd like to know how loud it is when in a system running with 5 or 6 120mm fans like a lot of high end cases have these days.
Associate
You can find the vids here. Also, one of the guys over at EVGA forums did a test with some screens. Here's the link
Last edited:
Associate
from the EVGA forum linked by KWJay:
the guy posting had a 5870. from my experience, 5870 at 70% is not bearable at all!
anything higher than 40% on 5870 is too much noise. but a stock reference 5870 can manage to stay under 70c in gaming with 40% and under 75c in Furmark.
the guy posting had a 5870. from my experience, 5870 at 70% is not bearable at all!
anything higher than 40% on 5870 is too much noise. but a stock reference 5870 can manage to stay under 70c in gaming with 40% and under 75c in Furmark.
Caporegime
- Joined
- 18 Oct 2002
- Posts
- 33,188
Exactly, 70% fan on ATI is ruddy awful, overclocked, overvolted my 5850 doesn't EVER go past 40% fan speed to maintain decent temps, I normally cap it out at 33% and you can hear the difference between idle of 24% and 36% where it starts to get not loud, but louder than I'd like, 50%+ is awful noise wise and 70% is ridiculous. My 4870x2 which ran WAY hotter and used a heck of a lot more juice never went above 45% either overclocked, 70% fan speed, would be unbearable for most people for day to day use.
Remember the few guys that have them on those forums are mostly big overclockers, most people are happy to run 100% fans for a few hours benchmarking a card, but won't run it that way 24/7.
As for noise, people will not buy a AMD or Nvidia card with the current type fans, that ran at 100% all the time, they would most certainly prefer higher temps and quieter than lower temps and silly noise.
Silly noise effects the end user, as long as temps are below throttling and after that death temps, it doesn't remotely effect the end user, and they wouldn't even be aware if they didn't have temp monitoring software which most people don't watch all day long. Noise you can't get away from, temps don't really matter.
My 5850 will hover around 80C while heavily clocked/volted and the fan still stays low, because 80C isn't deadly for a gpu, the noise would be worse than the temp for almost everyone.
Caporegime
Max out furmark with 1920x1080 MSAA, post FX, displacement mapping, extreme mode.
You'll need more than 40% fan..
I don't quite agree, high temps damage hardware, high noise is irritating, but wont break anything, except maybe your ears in this case
Obviously the preference is to have both low temps and noise, which we know can be done (the 5 series ati cards as example).
Saying that I agree nobody (normally) would want there fan running at 100% while using there pc.
I still think the only practical way to use the 480 is under water, but it doesn't bother me as I've no intention of buying one.
Still interesting to see how other users deal with these cards though.
Caporegime
- Joined
- 18 Oct 2002
- Posts
- 33,188
Yup, when I put my computer in the oven, and run solitare that would also make my gpu run hotter and fan go up.....
But I don't do that, and I don't "play" or benchmark Furmark, Furmark is a piece of crap and people who run it are just daft.
If not a single game can push temps as high as furmark, I don't care if it does push fans higher and temps up, if in the past 2 years Furmark hasn't replicated the conditions from any game, on any card, ever, its as pointless as my example of running your computer in an oven.
THere is NO game that has pushed my fan above 40% and actually I've yet to see it above 36% IN GAME, except where I've set the fan higher to see what the temp difference is on purpose.
AS to Spazzfish, yes high temps damage hardware, 80C is NOT high, its high for a human, its high for a plant, its NOT high for a chip DESIGNED TO RUN at that temperature.
The choice between a 100% fan and 70C temps, and 50% fan and 90C temps, when its designed not to break till 110C, the noise effects you the temp difference does not. If it was a choice between 100% fan and 70C< and 50% fan and 120C< and the card breaks at 110C, thats a different matter, neither the 5870 nor the 480gtx is in that situation.
Soldato
There is a big difference between having 90c and 70c blowing around in your case. It raises the ambient temp in the case to a higher point, meaning better air flow/fans are required. leaves less head room for temp increases in the summer. Running duel cards would increase the problem further.
I was also under the impression that heat reduces the life span of your hardware (where silicon is concerned).
Like i said before having both cool and quiet is preference, but when building a pc normally you make sure it runs cool before quiet.
Though with these cards i appreciate you are looking at noise beyond what is acceptable, at least to me.
The 480 runs at 94c stock when playing crysis. That is to close to its cut off which is about 105c.
In summer that card will be shutting itself down, especially when a bit of dust gets in the system.
Noise in this case is definitely preferable.
Last edited:
Soldato
- Joined
- 1 Apr 2010
- Posts
- 3,034
great that would be appreciated
I found that video for you.
http://www.youtube.com/watch?v=Sax8TvPDuYU&feature=player_embedded#
Edit: I see someone has beaten me to it.
Last edited:
Soldato
- Joined
- 1 Apr 2010
- Posts
- 3,034
Associate
- Joined
- 18 Jul 2009
- Posts
- 629
http://www.youtube.com/user/trubritar
Trub has now upload a few comparison benchmarks. Say what you like about him he does bench things always in the same way and on top of the line gear.
Trub has now upload a few comparison benchmarks. Say what you like about him he does bench things always in the same way and on top of the line gear.
Soldato
- Joined
- 6 Feb 2010
- Posts
- 14,594
Always posted this under a different topic, but I guess should bring it over here too:
Test kit: 3.2GHz Intel Core i7-965 Extreme Edition CPU, Asus P6T Deluxe, 6GB Corsair 1,600Mhz DDR3
Drivers: Nvidia Forceware 197.17 BETA WHQL, ATI Catalyst 10.3 WHQL
Battlefield: Bad Company 2 (DX 11, Win7 64-bit)
1,920x1,200, no AA, 16xAF
HD5970: min 71fps, average 116fps
GTX480: min 56fps, average 75fps
HD5870: min 52fps, average 74fps
1,920x1,200, 4xAA, 16xAF
HD5970: min 56fps, average 91fps
GTX480: min 44fps, average 66fps
HD5870: min 32fps, average 58fps
(source: Custom PC Issue 081 June 2010, p11)
Test kit: 3.2GHz Intel Core i7-965 Extreme Edition CPU, Asus P6T Deluxe, 6GB Corsair 1,600Mhz DDR3
Drivers: Nvidia Forceware 197.17 BETA WHQL, ATI Catalyst 10.3 WHQL
Battlefield: Bad Company 2 (DX 11, Win7 64-bit)
1,920x1,200, no AA, 16xAF
HD5970: min 71fps, average 116fps
GTX480: min 56fps, average 75fps
HD5870: min 52fps, average 74fps
1,920x1,200, 4xAA, 16xAF
HD5970: min 56fps, average 91fps
GTX480: min 44fps, average 66fps
HD5870: min 32fps, average 58fps
(source: Custom PC Issue 081 June 2010, p11)
Caporegime
- Joined
- 18 Oct 2002
- Posts
- 33,188
The problem is they've already had their dustbuster disaster and don't want to be known for having a dustbuster mark 2, though funily enough they canned, from what I recall anyway, the 5800 because it was too slow for the noise required and just wasn't very good and had the 5900 very close to ready so it was logical, but afaik had they wanted to they could have just made lots of them, yields weren't the issue, just noise, no one wanted a card that was that loud.
As for the 480gtx and fan speeds, they don't have to follow a specific ratio of 10% more fan every 10C increase, they'll keep it as low as possible till the last possible second, expect that it will infact hit 100% fan speed probably at 98-102C somewhere and I think that will turn a poor and overpriced 480gtx experience for the few, into a quite horrible experience as they hit hot weather. We'll see though, even worse in these situations is the actual fan noise changing, 100% speed the whole time is much less irritating than 100%, follow by a couple minutes at 70%, then 100% for a minute again. Even a loud noise you can get somewhat used to as long as its constant, the second it becomes a constantly changing noise you can't adjust or get used to it.
I tend to pick the max quiet fan speed I can get, in the 5850's case, about 34%, and stick it there in my gaming/overclocked profile, so even when it doesn't need quite that, I get used to the noise and don't get the changing pitch.
Silicon does degrade quicker the higher the temperature, but again it will be based on design, failure, or lifespan vs temp will be a exponential scale, the long slow part of the graph, well the throttling speed will be a little bit before it suddenly goes up dramatically.
If you go and check out the "how a cpu is made" type thread in the CPU section, you'll see CPU's as they are made go through heating at various temps over a 2 month period pretty much before they are ready, 365c for a short period at one point, and 140C for 12 hours to check nothing fails, they wouldn't last at all long at those temperatures over weeks/months but those are the kind of "failure" temps you're looking at to be honest. 105C would probably last a good year or more, 95c would last another year, and 85c would last another two years on top of that.
The biggest issue is those are general temps, some bits are likely to run hotter, not by too much though, some bits cooler. its the temps of power regulation circuits that are probably most in danger of early failure. CPU's/GPU's really are made to astonishing high levels of quality and will last a lot of abuse, cheap caps made by the millions in chinese factories....... less so.
Last edited: