LOL@Evga response to 8800GT fan issue

Table21 said:
Maybe its better for you guys in the states but every time that I RMA something it cost me a fortune.

RMAing a part is not a solution to a overheating problem. Could you imagine I a company like Western digital would say that too you. "Oh ya, don't worry about losing all your data and not having a working computer for 3 weeks...you get a new drive at the end of it"
Click to expand...

Thats a bad example there. Video cards and hard drives are two very different things. Just install rivatuner and set the speeds manually, /thread
 
Table21 said:
. . . RMAing a part is not a solution to a overheating problem. Could you imagine I a company like Western digital would say that too you. "Oh ya, don't worry about losing all your data and not having a working computer for 3 weeks...you get a new drive at the end of it"
Click to expand...

RMA it now for overheating or later for fan failure, your choice. :(

Also, a good manufacturer would offer a cross-ship option so you aren't without a card. some even cover shipping costs if the RMA is due to defect. Worth checking on before you buy brand X.
 
RaidalG said:
....some even cover shipping costs if the RMA is due to defect. Worth checking on before you buy brand X.
Click to expand...

Yes, people should also practice more responsibility about things like that. PC gaming isn't for everybody and it actually requires some knowledge and homework.
 
RaidalG said:
RMA it now for overheating or later for fan failure, your choice. :(
Click to expand...

Most [H] guys here, will replace the fan themselves, with something better, rather than RMA a card. I know I would. It's a cheap ass fan that is easily replaced, even better, replace with a better HSF assembly....
 
You are correct that time will tell. I will bet good money that a year from now if you look back, this card will be proven as the great design that it is and EVGA will not have massive returns.
Click to expand...

Perhaps, although I would sleep better at night if I just slap an after market cooler on it :). While I understand the point of this thread is that EVGA claims this is all ok. Perhaps it is. A real question though is "What does EVGA think the life expectancy of thier card is". Do they see this thing only having to last 1 year or 3 or 5? The old 9800 is what 5 or 6 years old (I'm not gonna bother to look it up) and will still run games these days (not at full settings, but you can more than play WoW with the graphics throttled back some).

Time will tell, but I'm to lazy to wait so aftermarket ftl :(
 
Table21 said:
Maybe its better for you guys in the states but every time that I RMA something it cost me a fortune.

For example I shoot a product from Canada over to the USA, it cost $20 bucks for shipping, then they send it back, using UPS/Fedex and I get charged duty/tax"
Click to expand...

Sometimes, but it also depends on who you're RMA'ing your item to. I'm from the US and I bought a HIS ICEQ3 X1950 Pro from OverclockersUK in the UK which turned out to be bad. Before you say it, it was actually cheaper than buying it in the US with a sale they had. Paid about $21.00 in shipping and when I had to RMA the card, they refunded me the shipping costs. Twice actually, since I received 2 bad cards due to a flaw in the Revision 1 cards.
 
Since the 8800Gt's are all the same card (except for that oddball new guy) would not the EVGA bios work on say, a Gigabyte? It should....
 
While theoretically it should, there have been instances where people have tried to flash XFX motherboards with EVGA bioses and it didnt turn out too well.
 
RussianHAXOR said:
While theoretically it should, there have been instances where people have tried to flash XFX motherboards with EVGA bioses and it didnt turn out too well.
Click to expand...

I was hoping to find a version of nibitor for the 8800GT's, apparently, the latest version can't read the bios....it "sees" the card but can't pull any info.
 
If you don't want to screw with your bios,see this information I posted from another thread. Basically the bios update is just going to do the same thing, except hardcode it. I recommend looking into this because you can adjust it more.

Nvidia's fan architechture is very different. You can set it to automatically spin up and down using the Low-level system/settings/Fan/Auto par of RivaTuner..

First you have to enable a registry key (from within Riva) to change this. This is where the 29% until 100c, then 100% thing is set.

Then you should read up on how the fan control works, it's a bit complicated, http://forums.guru3d.com/showthread.php?p=2500053

That thread explains it. Once you've figured it out, I've got some good settings for you:

Duty cycle min = 34
Duty cycle max = 100
T min = 56
T range = 8
T operating = 110
T low = 0
T high = 110

This basically sets it so the fan goes to 100% around 80c, otherwise it dynamically starts adjusting once temps are above 56c. The fan is set to run at a minimum of 34% all the time.

The tricky part is calculating the range, rather than having a system where u just say 100% at 80c, the fan uses the range to calculate how "dynamic" the fan is.

Ie., if you are running a game and use this method you will hear this thing occasionally speeding up and down a bit more than it would if you were using defined fan profiles...This doesn't bug me as it so happens most games create a constant temperature of around 64c with little variance so the fan doesn't change speed in an annoying fashion.

Using my settings, the fan doesn't need to ever go over 60% or so in most games, and my load temp is 64c or so. Crysis is the one exception, that game heats this card up, and the fan goes to 70% sometimes and I get around 70c...

this system is superior to fan profiles because the fan dynamically and immediately adjusts to temperature, so it's never too loud, never too soft, it's always just right!

BTW, my hunch for why we have the default 29% until 100c setting is because this fan control system is complicated. The manual (available in PDF through that thread) is like 100 pages long. There's a lot engineers can do to optimize this fan systems' cooling vs silence but it takes a lot of work. I'm sure Nvidia just wanted to get this thing to market so they gave us a noob setting.
 
jAkUp said:
EVGA has a new BIOS on their site that increases fanspeed :)
Click to expand...

Well at least it didnt take them too awfull long to figure it out.

I guess all the "Nvidia and EVGA can do no wrong" guys will say they are reacting to customer demand to provide a louder fan.
 
Just received my Inno3D 8800GT OC...

Default fan speed is 45% (2500 RPM) idle temp 47 degrees :confused:
 
Met-AL said:
OP spreading FUD around again?
Click to expand...

for once no. I've got one of these cards and have the same issue. I'm actually considering the 8800GTX step-up: It may take 2 slots, but it actually runs at stock speeds like it should and stays cool. with my fan at 80%, it's LOUD, and I can STILL cook an egg on the HSF should I choose.
 
Hazaro said:
I'll put in my experience. As soon as I got the 8800GT SC I fired up the Crysis demo and played for an hour at 92C load constant (stock speed). No problems at all.

Then I stuck on my Accelero S1 and got 52C load, and the card runs the same with same overclocks.

imo the stock cooler is fine and all you have to do is turn up the speed if you want to.

The eVGA rep is just saying to chill, but 80+C worries me :p
Click to expand...

I've had the opposite experience. Crysis locks up within the first 5 minuts of play with 80-85c on the card, unless I increase the fan speed to 80%. 80% keeps it cooler, but I have absolutely no overclocking headroom on the card, and it sounds like a leafblower or a small vacuum. Fan speed does not increase on it's own, and the card is unusable. I'm debating between RMAing to eVGA, or simply using step-up to get something with a real cooler on it. I'd rather not modify the card, and the noise is killing me.
 
lopoetve said:
I've had the opposite experience. Crysis locks up within the first 5 minutes of play with 80-85c on the card, unless I increase the fan speed to 80%. 80% keeps it cooler, but I have absolutely no overclocking headroom on the card, and it sounds like a leafblower or a small vacuum. Fan speed does not increase on it's own, and the card is unusable. I'm debating between RMAing to eVGA, or simply using step-up to get something with a real cooler on it. I'd rather not modify the card, and the noise is killing me.
Click to expand...

You really should consider RMA'ing the card as that is NOT normal. We have seen several posts from people here with cards running fine with the default fan settings. The lack of a widespread recall tells me their are lots of people with good cards. That eliminates the design as the issue. If its not the design, it must be manufacturing errors.
Unfortunately that does happen and it is most likely to happen with a new design. the bugs get worked out and the product is stable from them on.

I wouldn't accept paying approx. $300.00 for a video card that doesn't run flawlessly at stock settings and I don't think anyone else should. The fact that their are tools that allow you to tweak the fan to cool the card better shouldn't make you accept its flaws.

What people (and the OP) don't want to accept is that a chip can run at 100C and never have reliability issues. I can tell you from experience that their is silicon out there designed to run at much higher temperatures reliably for many years, and it does. We shouldn't be so quick to scoff at EVGA tech support for making the statements that started this whole thread as it is not unlikely that this is a design decision instead of a flaw.

Having said that, I can accept that 29% may well have been an error and maybe it should be higher, but instead of adjusting it ourselves, I think we should be RMA'ing the cards to make the manufacturer more quickly aware of the problem. If a couple hundred customers are asking for RMA's daily, you'll see quick action, be it a bios release, or recall.
 
RaidalG said:
You really should consider RMA'ing the card as that is NOT normal. We have seen several posts from people here with cards running fine with the default fan settings. The lack of a widespread recall tells me their are lots of people with good cards. That eliminates the design as the issue. If its not the design, it must be manufacturing errors.
Unfortunately that does happen and it is most likely to happen with a new design. the bugs get worked out and the product is stable from them on.

I wouldn't accept paying approx. $300.00 for a video card that doesn't run flawlessly at stock settings and I don't think anyone else should. The fact that their are tools that allow you to tweak the fan to cool the card better shouldn't make you accept its flaws.

What people (and the OP) don't want to accept is that a chip can run at 100C and never have reliability issues. I can tell you from experience that their is silicon out there designed to run at much higher temperatures reliably for many years, and it does. We shouldn't be so quick to scoff at EVGA tech support for making the statements that started this whole thread as it is not unlikely that this is a design decision instead of a flaw.

Having said that, I can accept that 29% may well have been an error and maybe it should be higher, but instead of adjusting it ourselves, I think we should be RMA'ing the cards to make the manufacturer more quickly aware of the problem. If a couple hundred customers are asking for RMA's daily, you'll see quick action, be it a bios release, or recall.
Click to expand...

Yeah, I'm about to. Guess I'll find out how good eVGA's RMA program is. I paid for the 2day upgrade.
 
Liquid Cooling FTW.

One thing I keep seeing overlooked in this thread is that just because the core is at 100c does not mean that the rest of the components are at 100c anyways. Hell when I go to liquid all I'm planning to add to anny of the regulators and memory or other components are some ramsinks which should suffice. GPU will me liquid cooled because it is the hottest part.
 
Justintoxicated said:
Liquid Cooling FTW.

One thing I keep seeing overlooked in this thread is that just because the core is at 100c does not mean that the rest of the components are at 100c anyways. Hell when I go to liquid all I'm planning to add to anny of the regulators and memory or other components are some ramsinks which should suffice. GPU will me liquid cooled because it is the hottest part.
Click to expand...

That's a great point! That 100C is an internal temperature in the chip itself usually. The rest of the circuitry will never see anything close to those temps.
 
The fact that there is a bios update released means that someone at EVGA is being pushed around by customer opinions or there really was a problem. Neither of those are encouraging. IF the technical people are being pushed by customers or they released the fan speeds wrong it scares me ; ;.
 
I want to barry a comment i have in this thread:

Bencharmarks aren't telling the whole story about experience quality and i'm focusing on a certain thing, studdering. Many people have noticed that when you try and look from side to side at a large angle the game has a very brief studder. This is really annoying. I ran benchmarks on my GT SSC and my GTX and the GT actually beat out the GTX. However, when I had the GTX installed it NEVER studdered while playing Crysis. Before i even put in the order for the GT I was noticing how smooth the GTX played the game, as far as zero studders. The GT studders.. and I kinda wonder why and also why there are no benchmarks that seem to capture this fenominon that has a large bearing on experience quality.
 
larkin said:
I want to barry a comment i have in this thread:

Bencharmarks aren't telling the whole story about experience quality and i'm focusing on a certain thing, studdering. Many people have noticed that when you try and look from side to side at a large angle the game has a very brief studder. This is really annoying. I ran benchmarks on my GT SSC and my GTX and the GT actually beat out the GTX. However, when I had the GTX installed it NEVER studdered while playing Crysis. Before i even put in the order for the GT I was noticing how smooth the GTX played the game, as far as zero studders. The GT studders.. and I kinda wonder why and also why there are no benchmarks that seem to capture this fenominon that has a large bearing on experience quality.
Click to expand...

a single second of low frames doesn't effect a large sampling, like a benchmark takes. We're talking about a change of MAYBE 1-2fps on the average because of stuttering.
 
I don't think he was asking why the NORMAL benchmarks don't capture is as much as he is asking why we use benchmarks that clearly represent quality game play. I would easily trade 1-2 FPS for no stuttering. It is really destroys my suspension of disbelief when playing a game. I am not sure there is an easy fix however. You could probably write an algorithm that would show this. Something like:
(AverageFPS)-(SomeConstant)((Time Between frames)^2/(total number of frames)). But this would be a bit more complicated as well as confusing. Trying to get some standard implemented would be hard if not impossible. How many of us would like some number on the side of the box that tells us how fast a card really is? Take my X850 XT (pro unlocked to the XT) and compare it to my wife's 7900. Which is faster depends on AA levels, game we are playing, and a number of other factors. I think this is what is largely hurting the PC gaming industry vs say a Xbox or PS3 where you know if you buy a game it will play.

Hmm..... Thread jacking at its worst.... Please forgive me >.<
 
Have you guys never looked at a HardOCP video card review?

This is exactly what they do. They find the settings for each card that give the most acceptable comprimise between playability and graphics and then compare those settings instead of raw FPS.

Then they give you the FPS over time graph for the entire benchmark for each card, all superimposed.
 
vengence said:
I don't think he was asking why the NORMAL benchmarks don't capture is as much as he is asking why we use benchmarks that clearly represent quality game play. I would easily trade 1-2 FPS for no stuttering. It is really destroys my suspension of disbelief when playing a game. I am not sure there is an easy fix however. You could probably write an algorithm that would show this. Something like:
(AverageFPS)-(SomeConstant)((Time Between frames)^2/(total number of frames)). But this would be a bit more complicated as well as confusing. Trying to get some standard implemented would be hard if not impossible. How many of us would like some number on the side of the box that tells us how fast a card really is? Take my X850 XT (pro unlocked to the XT) and compare it to my wife's 7900. Which is faster depends on AA levels, game we are playing, and a number of other factors. I think this is what is largely hurting the PC gaming industry vs say a Xbox or PS3 where you know if you buy a game it will play.

Hmm..... Thread jacking at its worst.... Please forgive me >.<
Click to expand...

To put it simply, I expect the resolution on the graphs isn't good enough ;) the dip is there, it's just too small to see.
 
I am loving this card, I am seeing idle temps around 52c, I have note really checked what they are getting to under load though. I am assuming this is the norm for these cards as I had an aftermarket cooler on my old 7950GT and temps were in the low 40s idle!!!
 
Mine sits @ like 49c.

Ive been running it at 45% fan speed in riva tuner so far today though.

Its fun to jack it up to 100% fan speed and listen to the vaccuum suck!
 
I installed a VF-900 on my XFX 8800GT. Idle temps dropped about 10*C give or take, and I don't think it goes above 60*C while gaming. I'm not anywhere near that system at the moment, so I can't give exact numbers. At any rate, it's much, much better than stock.
 
ClariorHincHonos said:
I installed a VF-900 on my XFX 8800GT. Idle temps dropped about 10*C give or take, and I don't think it goes above 60*C while gaming. I'm not anywhere near that system at the moment, so I can't give exact numbers. At any rate, it's much, much better than stock.
Click to expand...

I gave away a VF-900 as a donation to a LAN one time because I bought it for a card that I ended up getting rid of. Why oh why did I do such a thing....
 
8800GT refrence cooler redesigned

Well this hits the nail into coffin so to speak, as if you needed a crystal ball to tell that 8800GT reference cooler is inadequate for the job. Hopefully the new cooler keeps the temps low enough, still nasty for people who already bought 8800GT with the "old" cooler.
 
sam0t said:
8800GT refrence cooler redesigned

Well this hits the nail into coffin so to speak, as if you needed a crystal ball to tell that 8800GT reference cooler is inadequate for the job. Hopefully the new cooler keeps the temps low enough, still nasty for people who already bought 8800GT with the "old" cooler.
Click to expand...

Sadly its only 2 to 3C cooler, I dont think its enough of an improvement to be concerned.
I'd change the cooler anyway (I have done to mine), especially as the Accelero S1 is dirt cheap and awesome performance, 46C load fully clocked here.
Its far better for performance, temps and price to buy a cheaper card and fit one of these coolers than to get a highly clocked card with stock heatsink.
 
Sounds like Nvidia still doesn't think heat is the problem, but they addressed the noise issue. Those tests didn't measure dBs though. I would like to see those. I can't wait to get mine and get a nice aftermarket cooler.
 
You must log in or register to reply here.
Back
Top