Quote:

Originally Posted by 2tired

hi I have a r9 290 that gets really hot during gaming. Right now its idling at 40 but it shoots up to 90 while Im playing BF4. Air flow is weak in my case and I plan on adding some more fans, but will reduce my temps by a lot?

I read one dude who had the same problem. He switched to a 780 and never got temps above 70 while he was playing BF4. Can good airflow reduce my temps by 20 degrees? THis is my last ati card btw. Cant believe this is a concern with a $400 card.

Besides fans, is there anything else I can do to lower temps? IS this a common problem wioth all r9 290s?
My case is a nxzt phantom.

thanks

which r9 290 do you have

 

Its normal. The card is designed to run at 90C.

 

Its normal for reference cards anyway. I had a 480 before this so I know all about GPU temps and R9 290's are right up there.

 

like zealot said, its normal. But optimizing airflow may help a bit with temps

this could be helpful: http://www.overclock.net/t/1476618/water-to-big-air-question#post_22020875

 

Quote:

Originally Posted by 2tired

hi I have a r9 290 that gets really hot during gaming. Right now its idling at 40 but it shoots up to 90 while Im playing BF4. Air flow is weak in my case and I plan on adding some more fans, but will reduce my temps by a lot?

I read one dude who had the same problem. He switched to a 780 and never got temps above 70 while he was playing BF4. Can good airflow reduce my temps by 20 degrees? THis is my last ati card btw. Cant believe this is a concern with a $400 card.

Besides fans, is there anything else I can do to lower temps? IS this a common problem wioth all r9 290s?
My case is a nxzt phantom.

thanks

The 290-series have a thermal roof of 95c before it starts to throttle, it will clock itself lower and lower until conditions are met. Nvidia has decided to put their thermal limit at roughly 80c. There is nothing troubling with this. All the components used in the cards are rated and often garanteed to work atleast X amount of hours at temperatures way above that.

You can, if you're really, really, really concerned (which you should not be) about the temperature, lower the target temperature in the driver. Mind you, performance is affected by this. Or, you could ramp up the fan-speed and sacrifice silence instead. The only real and "noticable" advantage about lowering the temperature of the card without it throttling, is slightly lower power-usage, say, 5-15 watts. Also, note that temperature is NOT the same as heat. My watercooled 290 is dumping the same amout of heat into the room, and thus heating it just as fast as when it was air-cooled.

Adding fans may or may not lower the temperature. It depends on the cooler and the case-fan layout. For instance, the Asus DCII-cooler on the 290 performed like a champ on an open testbench, but when it was put in a case, it was among the worst.

Just a side-note, in case you are worried about this aswell. After you are done gaming, the card will still have a high core-temperature. That is because the fan works in a different way now, taking into account quite a lot of variables. When the card is not under load, it will spin down, and thus reduce the temperature of the core slower than going the traditional route regarding the fan.

 

Quote:

Originally Posted by Hattifnatten

The 290-series have a thermal roof of 95c before it starts to throttle, it will clock itself lower and lower until conditions are met. Nvidia has decided to put their thermal limit at roughly 80c. There is nothing troubling with this. All the components used in the cards are rated and often garanteed to work atleast X amount of hours at temperatures way above that.

You can, if you're really, really, really concerned (which you should not be) about the temperature, lower the target temperature in the driver. Mind you, performance is affected by this. Or, you could ramp up the fan-speed and sacrifice silence instead. The only real and "noticable" advantage about lowering the temperature of the card without it throttling, is slightly lower power-usage, say, 5-15 watts. Also, note that temperature is NOT the same as heat. My watercooled 290 is dumping the same amout of heat into the room, and thus heating it just as fast as when it was air-cooled.

Adding fans may or may not lower the temperature. It depends on the cooler and the case-fan layout. For instance, the Asus DCII-cooler on the 290 performed like a champ on an open testbench, but when it was put in a case, it was among the worst.

Just a side-note, in case you are worried about this aswell. After you are done gaming, the card will still have a high core-temperature. That is because the fan works in a different way now, taking into account quite a lot of variables. When the card is not under load, it will spin down, and thus reduce the temperature of the core slower than going the traditional route regarding the fan.

man this post is spot on. I did notice a difference with MSI afterburner on. I lowered it to about 70 degrees, but I always noticed the game ran much better with out.

So you are guys saying 90 degrees is ok? Man i was pooping my pants over here scared to death lol. 90 degrees seemed really high for a gpu and I didnt want it to be that high for an hour or whatever.

One quick question: do you any of you guys have AMD OVerdrive enabled? Is that useful? I turned mine off after having it on for a while.

you guys are the best. Thanks for clearing all this up.

 

Quote:

Originally Posted by 2tired

man this post is spot on. I did notice a difference with MSI afterburner on. I lowered it to about 70 degrees, but I always noticed the game ran much better with out.

So you are guys saying 90 degrees is ok? Man i was pooping my pants over here scared to death lol. 90 degrees seemed really high for a gpu and I didnt want it to be that high for an hour or whatever.

One quick question: do you any of you guys have AMD OVerdrive enabled? Is that useful? I turned mine off after having it on for a while.

you guys are the best. Thanks for clearing all this up.

The thing is that older GPUs if you had 90C+ you really did not know if in a HOT day or a very demanding game you might hit even higher. 290 cards have a wall 94C. At that point the card will slow down to keep itself safe. Doing so you will get lower performance. AMD Overdrive is not needed. Keep it off. Play with MSI AB. If you don't mind noise you can make a more aggressive fan profile to keep the card running full speed at lower temps but run louder.

 

I would try a gelid icy vision with vrm`s kit or a kraken g10 with a better clc than a h50

 

Crank the fan if you don't mind noise, stock my ref 290 would chill around 70c with an aggressive fan profile via MSI AB.

 

How come the stock fan's don't start cranking up under load, when does it actually ever kick in without resorting to manually applying MSI? I just bought a Gigabyte radeon reference r9 290.. Also when i crank the stock fans over say 70% it creates more heat than it actually dissipates

 

Quote:

Originally Posted by Dothrevenant

How come the stock fan's don't start cranking up under load, when does it actually ever kick in without resorting to manually applying MSI? I just bought a Gigabyte radeon reference r9 290..

By default the fan speed will never rise above 40% (55% uber mode) unless the card approaches 94C.
Quote:

Originally Posted by Dothrevenant

Also when i crank the stock fans over say 70% it creates more heat than it actually dissipates

The fan doesn't draw enough power for this to be remotely possible, unless something is broken.

 

If it's a ref. 290, it's going to run hot period. nbd. This is my Tri-X after playing BF3 for ~3 hours (overclocked):

 

Quote:

Originally Posted by Dothrevenant

How come the stock fan's don't start cranking up under load, when does it actually ever kick in without resorting to manually applying MSI?

It's similar to what I said about the fan spinning down right after the card's been under load, even though the temperature is high. It's just the other way around, it will ramp up it's speed when the card approaches the temp target

 

I have another question

I was looking online and in this one youtube video this dude had a gtx 670 that never went above like 50 degrees in bf4. Would that gtx 670 have better performance than my r9 290 running at 90 degrees? I figured the cooler card would run better, right? The dudes hit reg was INSANE in that video and Im attributing it to video card temps (correct me if im wrong). IF thats the case then I will have to get a cooler for this card. I understand it can be anything, but I know my hardware is good enough for bf4. I just feel like GPU temps are holding me back

 

Quote:

Originally Posted by 2tired

I have another question

I was looking online and in this one youtube video this dude had a gtx 670 that never went above like 50 degrees in bf4. Would that gtx 670 have better performance than my r9 290 running at 90 degrees? I figured the cooler card would run better, right? The dudes hit reg was INSANE in that video and Im attributing it to video card temps (correct me if im wrong). IF thats the case then I will have to get a cooler for this card. I understand it can be anything, but I know my hardware is good enough for bf4. I just feel like GPU temps are holding me back

You can only really compare cards like that if they are identical, or atleast the same GPU. In a hypothetical situation, with multiple different, say, 290's and a very high ambient temperature to make sure the cards throttle as much as possible, then yes, the card with the best coolingcapabilities willl perform better because it does not throttle as much as the others.

Example:
40c ambient
Card 1, worst cooler: 95c, 753mhz
Card 2, better cooler: 95c, 802mhz
Card 3, even better: 91c, 947mhz

Here, card number 3 would obviously perform better than the rest, because it was able to prevent the gpu from thermal-throttling. But this is a absolute worst-case scenario, and I have yet to see even a reference 290 in a warm room throttle as much as the cards in the example. It was just to make the difference clear

The 670 is a "mid-tier" 1st gen Kepler card, and will be anihaleted by the top-tier cards from both sides. It uses a much smaller die, and even has cores lasered off. It's not the full GK104, which gets destroyed by GK110/Hawaii.
To illustrate the difference here, I'll use cpus. Imagine a dual-core Haswell. It uses less power, and is therefore easier to cool than a quad-core Haswell, which results in lower temperatures. But the quad-core got twice as many cores, and therefore, hypothetically, twice the performance, even though it runs a little hotter.

And if you're wondering, the guy in the video was watercooling his card to keep it at such temperatures

If you feel your card is temp-restricted, and therefore does not allow overclocking or something like that, you can look into aftermarket coolers. The Accelero Xtreme performs exceptionally well considering it's whispering quiet (

)

 

Quote:

Originally Posted by Hattifnatten

You can only really compare cards like that if they are identical, or atleast the same GPU. In a hypothetical situation, with multiple different, say, 290's and a very high ambient temperature to make sure the cards throttle as much as possible, then yes, the card with the best coolingcapabilities willl perform better because it does not throttle as much as the others.

Example:
40c ambient
Card 1, worst cooler: 95c, 753mhz
Card 2, better cooler: 95c, 802mhz
Card 3, even better: 91c, 947mhz

Here, card number 3 would obviously perform better than the rest, because it was able to prevent the gpu from thermal-throttling. But this is a absolute worst-case scenario, and I have yet to see even a reference 290 in a warm room throttle as much as the cards in the example. It was just to make the difference clear

The 670 is a "mid-tier" 1st gen Kepler card, and will be anihaleted by the top-tier cards from both sides. It uses a much smaller die, and even has cores lasered off. It's not the full GK104, which gets destroyed by GK110/Hawaii.
To illustrate the difference here, I'll use cpus. Imagine a dual-core Haswell. It uses less power, and is therefore easier to cool than a quad-core Haswell, which results in lower temperatures. But the quad-core got twice as many cores, and therefore, hypothetically, twice the performance, even though it runs a little hotter.

And if you're wondering, the guy in the video was watercooling his card to keep it at such temperatures

If you feel your card is temp-restricted, and therefore does not allow overclocking or something like that, you can look into aftermarket coolers. The Accelero Xtreme performs exceptionally well considering it's whispering quiet (

)

so lets say a GTX 780 70 degrees vs the r9 290 running at 90 degrees, which one would run better?

I think nvidia gpus are btter for fps games. The reason I say this is because Ive noticed a lot of high kd players run with nvidia. Dudes like with 3/4 kds, etc. Ive seen their youtube clips and their hit markers are sooo much better.

if I reduced my temps with a new cooler on my r9 290, do you think the bf4 would run better? im trying to figure this out

thanks

 

They perform the same. At launch the 290 (non-x) was faster than the 780 by about 1-2%.

Nvidia does have 2/3 of the descrete gpu market. You are therefore, of course bound to run into more nvidia-users. You're probably also experiencing some phsycological effect aswell. You see both cases, but only notice one. Have you noticed how it seems that every time you learn a new fancy word, it seems like everyone is using it? I'm guessing something similar is at work here.

Personal skill in games have nothing to do with the hardware they're playing on. There are many good players that have machines that can barely run the games (from my point of view). Your skill will not change wheter you're playing in 30, 60 or 120 fps. It is more pleasent though, to run at a higher framerate.

To find out if games will run better on the card if you slap on a better cooler, you must first find out if the card throttles. You can use Afterburner for this. Stock clocks fir a reference 290 is 947mhz on the core. Note that using vsync may cause the card to run at a lower clock, because it's not needed to run it higher to reach 60.
HOWEVER, the throttle temp for the 290's are 95c. Your card is at 90c, which means it does not throttle, which again means the framerate will be the same.

The only reason to change the cooler, is if you can't stand the noise from the reference one.

 

Quote:

Originally Posted by 2tired

so lets say a GTX 780 70 degrees vs the r9 290 running at 90 degrees, which one would run better?

I think nvidia gpus are btter for fps games. The reason I say this is because Ive noticed a lot of high kd players run with nvidia. Dudes like with 3/4 kds, etc. Ive seen their youtube clips and their hit markers are sooo much better.

if I reduced my temps with a new cooler on my r9 290, do you think the bf4 would run better? im trying to figure this out

thanks

Well I can tell you from experience ,the 780/780Ti is not better than r9 290/x cards.I am not going to comment on some of the stuff you posted over the last couple days.

I can try to help you out with your temps a little if you want to read what I am going to post.

I am running R9 290 Crossfire on air with reference and Non-Reference card in a Nvidia Haf-X case.

Reference card on the bottom,case has 1x200mm side intake,1x240mm front intake,2x200mm top exhaust and 120mm to exhaust air out the back.

Adding more fans work good,but I suggest custom fan profile is a good way to go.
Here is what temperatures look like with small overclock on the cards.

Crysis 3 runs so cards are working hard.No Throttling or other stuff.It just works.

R9 290 Crossfire Fans 100%


R9 290 Crossfire Custom Fan Profile


R9 290 Crossfire Auto Fan profile


There are several third party solution you can use for custom fan profile and voltage control of VCCD.

Here are a couple that are very good to use with voltage control .200mv and higher for high stable overclocks.

Sapphire Trixx
Asus GPU Tweak

Most Popular with voltage control as is .100mv for good overclock and fan control and much more
MSI Afterburner 3.0.1

Get the most out of your R9 290 card with tweaking
RadeonPro BETA (Automating 3D Settings)

You can also run a stable card without any voltage tweaking just fine and not have any throttling at all,and be cool also.

Example would be something like this with MSI after burner.

 

So you mean to tell that if I add LN2 cooling to my R9 290 I'll be able to get 50/1 K/D ratios?

Your card has nothing to do with your performance in games. Changing from Nvidia to AMD or AMD to Nvidia will not make you magically play better.
I have a laptop that never hits 70ºC on an Nvidia GPU and my K/D ratio on it are worst... maybe I need to get the GPU hotter. Probably gonna try n run it without a heatsink lmao

Or maybe I get a worst K/D ratio because I'm playing on a Full HD 13.3" inch display at the same distance as I usually play on my 22" inch screen lol. I feel like a blind bat inside BF4 on my laptop.

You have a good card, don't waste your time with imaginary gameplay improvements from changing cards/chaingng temperature.
Every R9 290 will turbo untill it hits 95ºC at 95ºC it will try to give the most performance. Card will crash after 100ºC are hit in case throttle isn't helping.
My GTX765M crashes at arround 85ºC. My old 6850 crashed at 75ºC, load temp for it was 68ºC overclocked.
You can't compare them, they are completely different.
Nvidia GTX480 would hit 95ºC.

Heat/Temperature is not something you can use to measure performance.

 

Quote:

Originally Posted by gerardfraser

You can also run a stable card without any voltage tweaking just fine and not have any throttling at all,and be cool also.

A blanket statement that is far from universally true.
Quote:

Originally Posted by Tokkan

Your card has nothing to do with your performance in games.

Frame rate, as well as resolution and certain other IQ settings can certainly influence one's effectiveness in certain titles.
Quote:

Originally Posted by Tokkan

You can't compare them, they are completely different.

Heat/Temperature is not something you can use to measure performance.

You can certainly compare the performance you can get out of a card at a given noise level.