Fixed: Yeah, not happy with this HD6950 (excessive heat with multimonitor problem)

Page 8 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Status
Not open for further replies.

discy

Junior Member
Jan 3, 2011
9
0
66
Well that's not true, Intel has the biggest market share in graphics and yet, with their twitchy driver support, had less issues than nVidia.

The point is it influences the result. You can't say nvidia = 28.8%, ati = 9.3% so ati is better . Same with Intel.
 
Last edited:

Nintendesert

Diamond Member
Mar 28, 2010
7,761
5
0
The point is it influences the result. You can't say nvidia = 28.8%, ati = 9.3% so ati is better . Same with Intel.


Of course it influences the result, Intel has a huge market share compared to Nvidia but yet caused a much smaller % of the crashes.

I'm a little confused now as to what you're arguing. Microsoft literally has a 100% market share for Vista and still accounted for a smaller % of the crashes.

No one was arguing that ATI was better but merely both companies continue to have problems and have a long track record of having problems. Neither one is infallible so for anyone to claim anything of the sort is disingenuous.
 

SirGCal

Member
May 11, 2005
122
1
0
www.sirgcal.com
Well, I've confirmed your driver install problems. You must use a specific version at the moment. I suspect the next actual CCC Release (10.13 or 12hotfix) will have this problem fixed. However asside from that I've had no problems. Even forcing the fan slow and overloading the GPU, it never hit 90 deg C. And never had a problem during a 24 hour stress test. All in all, it runs cool, quiet and fast. I can only suspect, as I did before, that the card you have is a dud. Perhaps a poor manufacturer or just a fluke of the manufacturing process...

Anyhow, I finished my full review about the transfer from the 4890 to the 6970 with benchmarks and pretty charts and graphs: http://www.sirgcal.com/blog/index.php?itemid=393
 

SirGCal

Member
May 11, 2005
122
1
0
www.sirgcal.com
The point is it influences the result. You can't say nvidia = 28.8%, ati = 9.3% so ati is better . Same with Intel.

That's not the point entirely of that entire post I made. Perhaps you should give it another read. I think Nintendesert understands what I was trying to convey.
 

badb0y

Diamond Member
Feb 22, 2010
4,015
30
91
So, WelshBloke, my experience with this card has been way better than yours. IMO, it's quieter than the EVGA GTX 460 SC EE and performance is awesome especially since I unlocked it to a 6970. I have the XFX model, my single-monitor idle is 31 degrees Celsius and fan is at 24%.

Temporary case ATM is a Antec DF-10 until I find a suitable Full Tower.

EDIT: Also I tried to stress test it with Kombuster and the temperature went up to about 80 degrees Celsius but my fan speed oddly didn't ramp up past 35%. Is this normal? Or should be worried this card might kill itself.
 
Last edited:

WelshBloke

Lifer
Jan 12, 2005
30,955
8,681
136
So, WelshBloke, my experience with this card has been way better than yours. IMO, it's quieter than the EVGA GTX 460 SC EE and performance is awesome especially since I unlocked it to a 6970. I have the XFX model, my single-monitor idle is 31 degrees Celsius and fan is at 24%.

Temporary case ATM is a Antec DF-10 until I find a suitable Full Tower.

EDIT: Also I tried to stress test it with Kombuster and the temperature went up to about 80 degrees Celsius but my fan speed oddly didn't ramp up past 35%. Is this normal? Or should be worried this card might kill itself.

The main problem I'm having is not with a single monitor. It was quiet and cool when it was in the shipping box as well but thats just as irrelevant.

My only problems with it are:


  • Runs hot with a secondary monitor of a different res attached.
  • Fan is annoying at anything over an idle. (>30% annoying, >40%WTF)
  • CCC is absolutely awful. I cant stress how bad it is, how a company as big as AMD cant make something so simple work better I dont understand (talking about the ergonomics of the control panel not the driver, the driver seems fine)
 

WelshBloke

Lifer
Jan 12, 2005
30,955
8,681
136
Well, I've confirmed your driver install problems. You must use a specific version at the moment. I suspect the next actual CCC Release (10.13 or 12hotfix) will have this problem fixed. However asside from that I've had no problems. Even forcing the fan slow and overloading the GPU, it never hit 90 deg C. And never had a problem during a 24 hour stress test. All in all, it runs cool, quiet and fast. I can only suspect, as I did before, that the card you have is a dud. Perhaps a poor manufacturer or just a fluke of the manufacturing process...

Anyhow, I finished my full review about the transfer from the 4890 to the 6970 with benchmarks and pretty charts and graphs: http://www.sirgcal.com/blog/index.php?itemid=393

Several other people in the thread have reported the exact same problem as me.
 

Jacky60

Golden Member
Jan 3, 2010
1,123
0
0
These kind of posts dont surprise me. I had my hell for almost 4 years with ati x800 xt pe

The hell is not in the hardware but in the buggy software. ATI cant make drivers and they never could even with the all in wonder you get desktop corruption. CCC installs sometimes other times it doesnt,, just a bi*ch. One driver your card is fast other driver things are slow slow.. I never will trust AMD ATI ,, Grab a nVidia card next time and you wont have these problems. Other then that RMA the card. gl

Strange because my X800XTPE never encountered any driver problems in line with the other owners here. Are you sure you installed the drivers correctly tweakboy?
 
Last edited:

betasub

Platinum Member
Mar 22, 2006
2,677
0
0
It was quiet and cool when it was in the shipping box as well but thats just as irrelevant.

No, that's facetious.

I agree that you have an issue when in multimonitor mode. I experience this too with my PowerColor 6950, although my temps are slightly lower, and my fan at 30% does not annoy me (either the fan noises are different, or our sensitivity to it differs).

Thanks for putting your work-around in the OP.
 

WelshBloke

Lifer
Jan 12, 2005
30,955
8,681
136
No, that's facetious.

Not really, I have run my computer with that monitor set up for years now and I wont ever be running a single monitor setup so how it runs with a single monitor is as relevant to me as how well it runs while still in the box.

I agree that you have an issue when in multimonitor mode. I experience this too with my PowerColor 6950, although my temps are slightly lower, and my fan at 30% does not annoy me (either the fan noises are different, or our sensitivity to it differs).

Thanks for putting your work-around in the OP.

Yeah, I've said lots of times I'm pretty unreasonable about computer noise.

Out of interest what level do you find the fan gets annoying at.

Hope the work around works ok for you, should get better when Afterburner gets voltage support then we can undervolt as well as underclock at idle. Might get multimonitor temps down to single monitor ones then as well.

I've been looking into aftermarket coolers as well but am having trouble finding out if Thermaltakes VRM coolers fit.
 

SirGCal

Member
May 11, 2005
122
1
0
www.sirgcal.com
Several other people in the thread have reported the exact same problem as me.

edited OP to clarify problem and show workaround
OK, I'll try to clarify even more... Does this help?

Fits fine in the case.

First problem the new CCC wont install (demo/preview whatever), and the old CCC is an ergonomic nightmare, its worse than the one I remember from years ago.

Second problem, it idles at 66C on the desktop and ramps up to stupid temperatures and levels of noise as soon as a game is launched, before shutting down. Temps above 95C, and before you ask my case is fine foe air cooling. Besides I've not even had time to put the sides back on yet.

Third problem, random programs look like ass. Window live mail text is bizarrely blurry but firefox is ok. :\

I've not had chance to try any video yet but this puppy is going back to the kennel, hopefully the replacement might be OK.

Edit: I'm going to bed.

Edit for anyone not reading the thread: This problem is due to running more than one monitor of differing resolutions. If you have low temps with 1 monitor attached thats fine and lovely but it has nothing to do with my situation.


My solution from later in the thread.


Well, I still suggest buying better products, especially if you plan to push them. Running more then one monitor obviously is going to consume more power and make it run hotter. But as a test, I have my 1920x1080 main and a 1600x900 second I pulled out of the spare room just to run a test. GPU is idling at mid 50's and running at 500MHz with the memory turned on (that's where your temps are coming from, but they shouldn't be that high). Fan speed is 27% and more silent than the ambient room noise (unable to measure with a DBMeter due to it not exceeding ambient levels, and this is in a private, quiet, over-insulated home if you were wondering.) and with an ambient temperature of 77 degrees F so it's quite warm in here to start with. So warm that the single-display ambient temperature at idle was mid 40's today. But the house is quite warm so... I haven't turned the AC on yet. Anyhow...

Here's the log.. See for yourself. I plugged in the second display at 2011-01-07 12:02:52.
http://www.sirgcal.com/misc/multi-res/GPU-Z_6970_multi_disp.txt



So if I cooled off the house to a proper ambient temperature, the idle temp for multi-res display would be in the low 50's to high 40's. The reason this is harder on cards is that it doesn't just do one larger display and split it in half or thirds like it would with resolution-matching monitors, it has to fully render two separate displays. So yea, obviously it's going to consume a bit more power. And actually, the longer I ran it, it started cooling off a bit (not in the chart) as the room cooled off, so the temperatures in that chart are stable operating temperatures.

So basically, unless your case is just that bad and your airflow is horrible or something, those temperatures are a bit high approaching 70C at idle. These are my temps with my case and the fans set to silent. So much so I can not hear them over the ambient hum (now granted, I do have a rather large 12TB server about 20 feet away behind me, I can hear it if I listen for it but...) The only noise I hear from my gaming rig when not gaming is the Raptor HDD when it's accessing... And that's with the side off even. If I put the side back on and get back to my negative pressure setup, temps will drop even more. It's always best to have a properly circulated case with the sides in position to ensure proper air routing and pressures. Plus colder ambient air is denser than warm air which is why it cools better, not just cause it's colder... FYI... (That's also why performance vehicles always try to get cold-air intakes so they basically get more air molecules in the same volume of atmosphere.)

So in summary, I still think your card is running too hot generally speaking and suspect a fault or ineffective manufacturer. But even the best manufacturer's put out the occasional lemon...

EDIT: Been running near two hours now with doing hardware video playback and a bit of games just to heat it up also. Still low to mid 50's in windows. Games still haven't hit 90...
 
Last edited:

WelshBloke

Lifer
Jan 12, 2005
30,955
8,681
136
I don't think the card gets hotter with two monitors because its doing more work, it gets hotter because its not dropping its clocks and voltages down to true 2d levels.
If it was getting hotter because of the extra work in driving more pixels temperature would scale with resolution, and older cards that dint have 2d/3d clocks didnt have such a temp change. Also if I force 3d clocks all the time I dont see a difference in temperature with 1 or 2 monitors.

I've changed the TIM and reseated the cooler and I idle at 61/62 and 30% fan (thats all at default CCC handling clocks and fan) you seem to idle at 57 so I'd say they were the same really.

I'm not sure your ambients would make a massive difference, unless they are quite extreme. The cooler seems to work best with a high airflow, but soon gets noisy.

I'm not too worried about my airflow, I've got a Silverstone tj09 and it seems pretty good for airflow (even with my fans on low) although I'd have thought that most cases that could fit an HD69x0 card would be ok.
 

SirGCal

Member
May 11, 2005
122
1
0
www.sirgcal.com
I don't think the card gets hotter with two monitors because its doing more work, it gets hotter because its not dropping its clocks and voltages down to true 2d levels.
If it was getting hotter because of the extra work in driving more pixels temperature would scale with resolution, and older cards that dint have 2d/3d clocks didnt have such a temp change. Also if I force 3d clocks all the time I dont see a difference in temperature with 1 or 2 monitors.

There's a huge difference in increased resolution and two independent output sources... If you don't understand driver programming, it is not easy to explain to you. But there is an enormous difference. It does however depend how the driver is coded to what it will effect.

I've changed the TIM and reseated the cooler and I idle at 61/62 and 30% fan (thats all at default CCC handling clocks and fan) you seem to idle at 57 so I'd say they were the same really.

If you live in an oven. I turned the air on an hour ago and ambient temp dropped to 73 and the idle temp also dropped to ~ 54-55. Ambient temperature does make an important difference, as well as actually where the computer is placed. If it's inside a desk, don't expect it to get good cool fresh air. Even just being in a corner can raise the temps a ton.

However, it does sound as though re-seating the cooler helped. Though it still shouldn't be that loud. I have the stock reference cooler style and at 28%, with it less than three feet from my face, I can barely hear it in perfect silence. With the case side on, I can not hear it at all. (and yes, I actually have phenomenal hearing.. More about that on my website if you really want to know). So perhaps your manufacturer used sub-quality fans or bearings... This is by far the quietest air-cooled gaming card I've ever had in reference form. Either way you still have an issue.

However, if you want to force your memories lower at the sake of some potential performance hickups, it's very easy to flash the bios of modern video cards with whatever custom settings you want... Change clocks to just about anything, fan speeds and trigger ramps, even voltages, etc. (Save the stock setup obviously to revert back at any time). It sounds like what you want doesn't exist. A high-performance card that doesn't run harder for huge replication areas... Comparing your card to others, it seems to be running quite warm and you complain about the audible levels. And plugging in the 2nd monitor at a separate resolution to my card only change the idle fan from 24 to 28% in a very hot ambient environment (I'm sweating just sitting here, well was before I turned the air on).

But even so, the card is in no danger. It is safe well up to over 100 deg C. So while it's obviously a sub-par quality build with a loud fan and poor cooler situation, it's still operating under spec. You make a card do twice as much work, and expect it to not use more power hence create more heat to do that work, that's a bit ludicrous to be honest regardless of how the drivers are setup. It's like asking a small car's engine to tow another car behind it but still get the same MPG as it did alone. Not going to happen. Even if it can successfully move twice the weight, it's going to take a huge hit in the energy consumption area as well as wear and tear.
 

blacklibrary

Junior Member
Jan 7, 2011
2
0
0
Just wanted to add to this:

I went through 3 separate HD 6950's due to the same issue Welsh is having. My card (unlocked) is idling at 68 C and I've seen it go up to 92 C with the default fan profile. It's a bit better when I crank the fan up to 40 or 50% but then it's pretty damn noisy. I'm also using a dual monitor setup with the SAME resolution on both.

There's also an annoying buzzing sound coming from it, which I've identified as the coil noise, apparently due to some compatibility issue with Corsair Tx750 power supply. I do not think this has anything to do with the temperature issue, which seems to be much more widely spread.

Is memory clock supposed to dial down or up automatically based on load? I was surprised to come across one particular screen shot that had memory profile at 250Mhz, and based on Welsh's solution, this would certainly improve the heat situation.

Edit: I haven't checked if I installed the preview version from the driver website (and naturally haven't installed the hotfix either). I'll be doing that later today and report my findings.
 
Last edited:

SirGCal

Member
May 11, 2005
122
1
0
www.sirgcal.com
There's also an annoying buzzing sound coming from it, which I've identified as the coil noise, apparently due to some compatibility issue with Corsair Tx750 power supply. I do not think this has anything to do with the temperature issue, which seems to be much more widely spread.

I doubt very much it has anything to do with that power supply. That's one stout unit and highly stable with minimal ripple which is why I run one myself. But a buzzing inductor or capacitor is a sure sign of questionable components.

But again, the temperature isn't really an issue... You want the card to do more, it's going to use more power. Which means more heat. Pretty simple logic. You can tweak the bios to a better profile for your particular intentions but... This is not AMD only either, you'll see this with all modern cards.

Again, this is just another example of why I always recommend buying from quality manufacturers. The budget lineup is just that and works as expected. Still your cards all seem fully functional just peachy, you just don't like their lower-quality fans, bearings and components (which cheap caps and such can also contribute to a huge heat increase situation, not necessarily just the memory and chip itself).

Again, I've installed now 5 of the 69xx series and not had any of these noise issues from any of the users (one who couldn't stand his 470 any longer due to it's noise, now loves the quietness of his system, who also is running multiple monitors (three now actually since he upgraded to the AMD chips again)). But I don't let them purchase the more budget oriented lineup unless they know what they're getting (some people are fine with noise and more heat, which means they're fine with those cards). And then sure, if you do have a problem, getting it resolved is often easier with the better manufacturers also. Some other budget manufacturers will trade you out a dozen times also cause it's cheaper for them to build the cheap units and offer some huge warranty and keep swapping out parts till it works and resell the returned items instead of spending more on higher quality parts and components to make a more tweaked lineup. In some cases you do pay for a name, but in many cases in the electronics world, you pay for higher bin-out and more quality parts.
 

badb0y

Diamond Member
Feb 22, 2010
4,015
30
91
To be quite honest guys, the AMD 6950 unlocked is making less noise for me than the GTX 460 SC EE did, I think it's more due to the EVGA's extremely aggressive fan profile that would jump to 56% as soon as the card hit 70 degrees but the 6950 unlocked is hovering around 35%~ fanspeed while temps on the GPU are at 78 degrees celsius. Overall I am extremely happy with the value I got, the performace of a 379$ card at the price of 260$ and if you factor in that I sold my GTX 460 for 140$, it comes to about a 120$ upgrade out of the pocket for me.
 

blacklibrary

Junior Member
Jan 7, 2011
2
0
0
I've tried installing the hotfix and was rewarded with higher min-cap on memory clock setting...what the hell? No improvement on the temp situation either...

I doubt very much it has anything to do with that power supply. That's one stout unit and highly stable with minimal ripple which is why I run one myself. But a buzzing inductor or capacitor is a sure sign of questionable components.
Maybe. I'm not too familiar with the problem myself, but I was assured that it's not harmful to the hardware (and it's hard to hear from outside the case anyways), so I stuck with the card...for now.

But again, the temperature isn't really an issue... You want the card to do more, it's going to use more power. Which means more heat. Pretty simple logic. You can tweak the bios to a better profile for your particular intentions but... This is not AMD only either, you'll see this with all modern cards.

Again, this is just another example of why I always recommend buying from quality manufacturers. The budget lineup is just that and works as expected. Still your cards all seem fully functional just peachy, you just don't like their lower-quality fans, bearings and components (which cheap caps and such can also contribute to a huge heat increase situation, not necessarily just the memory and chip itself).

Again, I've installed now 5 of the 69xx series and not had any of these noise issues from any of the users (one who couldn't stand his 470 any longer due to it's noise, now loves the quietness of his system, who also is running multiple monitors (three now actually since he upgraded to the AMD chips again)). But I don't let them purchase the more budget oriented lineup unless they know what they're getting (some people are fine with noise and more heat, which means they're fine with those cards). And then sure, if you do have a problem, getting it resolved is often easier with the better manufacturers also. Some other budget manufacturers will trade you out a dozen times also cause it's cheaper for them to build the cheap units and offer some huge warranty and keep swapping out parts till it works and resell the returned items instead of spending more on higher quality parts and components to make a more tweaked lineup. In some cases you do pay for a name, but in many cases in the electronics world, you pay for higher bin-out and more quality parts.

It makes sense that it'd use more power, hence more heat would be generated, but the important question is, is this a behaviour that would endanger the lifetime of your card/computer? I mean, from what I'm hearing from other forum posts, the graphic card that Welch and I have are pumping out about 20~30 C higher than what's been observed so far. It seems ridiculous that having a dual display could cause this kind of insane temperature difference.

Also, is 92 C really a safe temperature for graphic cards these days? This is my first replacement card since 3 years ago (gtx 9800), so I'm not too familiar with how the high end cards fair these days.

To add a bit more to this: I can use a custom fan profile that's only slightly more aggressive than the stock (30~60%, which usually remains between 30 and 45%) to keep it at around 70 C while at 30~50% GPU load, which is better than the 90's I got before, interestingly enough.
 
Last edited:

SirGCal

Member
May 11, 2005
122
1
0
www.sirgcal.com
It makes sense that it'd use more power, hence more heat would be generated, but the important question is, is this a behaviour that would endanger the lifetime of your card/computer? I mean, from what I'm hearing from other forum posts, the graphic card that Welch and I have are pumping out about 20~30 C higher than what's been observed so far. It seems ridiculous that having a dual display could cause this kind of insane temperature difference.

All electronics do have a limited lifespan in reality. How many electrons they can flow and how well decrease over time... How stressed they are also makes that decrease faster... The question here is; would you even care in the difference? Most likely no. So in proper operation single-monitor setup the card might last 15 years without a failure, when stressed with duals, maybe that lowers to 10? Still for a gaming rig, WELL below the upgrade threshold... (and no, I pulled those numbers right out of my huge fat backside... But the point is, unless you're making extremely long-term servers of some type, most home users won't be concerned with the lowered life due to normal wear/tear even stressed up using multi-monitors in this case).

Also, is 92 C really a safe temperature for graphic cards these days? This is my first replacement card since 3 years ago (gtx 9800), so I'm not too familiar with how the high end cards fair these days.
Depends entirely on the card. Even just a handful of years ago, that would have killed most cards outright. But lately they've been much more capable. Many cards even get far too hot to touch. Some high-performance cards of recent can hit 120C and still not fail. It's not recommended but... Furthermore, the 6900 series AMD cards have built-in safeguards to help prevent damage. When it sees itself using too much power, it will down-clock itself to protect the hardware.

To add a bit more to this: I can use a custom fan profile that's only slightly more aggressive than the stock (30~60%, which usually remains between 30 and 45%) to keep it at around 70 C while at 30~50% GPU load, which is better than the 90's I got before, interestingly enough.

One of the biggest short-falls of the budget builders are their fan profiles and such along with their cheaper parts. Getting a proper fan profile isn't that hard, just requires re-flashing the bios. But this is not for the uneducated (a bad flash can brick your card, and that's generally not covered under normal wear and tear warranties). However today this is a relatively safe operation with minimal education on the subject and you can get more aggressive, yet fully automatic fan profiles. Also many software packages today let you emulate the same thing (but those changes do not follow the card when moved and rely on the software, the only downside).

So in summary; most likely the buzzing cap or coil is not a problem, just an annoyance (and sign of cheap components). If it is going to fail, most likely it will do so relatively early in life (30-90 days). There are also ways to shut up the culprit if you can find it. As for life of the card, I really wouldn't worry about that unless you never ever plan to upgrade again. As long as it's in the safe operating range, you should be fine for an expected lifespan of 3-6 years depending on the manufacturer (more realistically, 7-12 years real-world use). There's always exceptions, and if you're setup is currently overstressed, it could be shorter, but if you're not seeing artifacts, crashes, etc, then odds are you're not in any real danger. If you are though then you may be teetering on the edge. It's really hard to say with any accuracy without seeing the system's problems first hand.
 

WelshBloke

Lifer
Jan 12, 2005
30,955
8,681
136
I doubt very much it has anything to do with that power supply. That's one stout unit and highly stable with minimal ripple which is why I run one myself. But a buzzing inductor or capacitor is a sure sign of questionable components.

But again, the temperature isn't really an issue... You want the card to do more, it's going to use more power. Which means more heat. Pretty simple logic. You can tweak the bios to a better profile for your particular intentions but... This is not AMD only either, you'll see this with all modern cards.

Again, this is just another example of why I always recommend buying from quality manufacturers. The budget lineup is just that and works as expected. Still your cards all seem fully functional just peachy, you just don't like their lower-quality fans, bearings and components (which cheap caps and such can also contribute to a huge heat increase situation, not necessarily just the memory and chip itself).

Again, I've installed now 5 of the 69xx series and not had any of these noise issues from any of the users (one who couldn't stand his 470 any longer due to it's noise, now loves the quietness of his system, who also is running multiple monitors (three now actually since he upgraded to the AMD chips again)). But I don't let them purchase the more budget oriented lineup unless they know what they're getting (some people are fine with noise and more heat, which means they're fine with those cards). And then sure, if you do have a problem, getting it resolved is often easier with the better manufacturers also. Some other budget manufacturers will trade you out a dozen times also cause it's cheaper for them to build the cheap units and offer some huge warranty and keep swapping out parts till it works and resell the returned items instead of spending more on higher quality parts and components to make a more tweaked lineup. In some cases you do pay for a name, but in many cases in the electronics world, you pay for higher bin-out and more quality parts.

They are all reference cards, the are identical apart from the stickers (and Asus flashes a custom bios). If you're suggesting that somehow the sticker has any bearing on the quality of the component...

And its not increased workload thats causing the heat its the fact that the card doesn't downclock or drop its voltages. If my laptop with Intel integrated graphics can push 2 monitors fine then the increased workload for an HD6950 is negligible.

Fact is the issue happens and it happens with cards across vendors, you blaming 'cheap' components is not helping.
 

WelshBloke

Lifer
Jan 12, 2005
30,955
8,681
136
Just wanted to add to this:

I went through 3 separate HD 6950's due to the same issue Welsh is having. My card (unlocked) is idling at 68 C and I've seen it go up to 92 C with the default fan profile. It's a bit better when I crank the fan up to 40 or 50% but then it's pretty damn noisy. I'm also using a dual monitor setup with the SAME resolution on both.

There's also an annoying buzzing sound coming from it, which I've identified as the coil noise, apparently due to some compatibility issue with Corsair Tx750 power supply. I do not think this has anything to do with the temperature issue, which seems to be much more widely spread.

Is memory clock supposed to dial down or up automatically based on load? I was surprised to come across one particular screen shot that had memory profile at 250Mhz, and based on Welsh's solution, this would certainly improve the heat situation.

Edit: I haven't checked if I installed the preview version from the driver website (and naturally haven't installed the hotfix either). I'll be doing that later today and report my findings.


Did you try the clockfix workaround in the OP?

Also enabling vsync can help with coilwhine sometimes, is it intermittent or there all the time?
 

SirGCal

Member
May 11, 2005
122
1
0
www.sirgcal.com
They are all reference cards, the are identical apart from the stickers (and Asus flashes a custom bios). If you're suggesting that somehow the sticker has any bearing on the quality of the component...

And its not increased workload thats causing the heat its the fact that the card doesn't downclock or drop its voltages. If my laptop with Intel integrated graphics can push 2 monitors fine then the increased workload for an HD6950 is negligible.

Fact is the issue happens and it happens with cards across vendors, you blaming 'cheap' components is not helping.

Ohhhh NO!.. Absolutely not... Reference just means design. NOT component usage. It says they will have a specific farad capacitor here, ohms of resistor there.. etc. NOT exactly which exact capacitor, resistor, etc to use for the job. Or which bearings to use in their fans. That's a HUGE amount of difference! Actually, they can not force the manufacturers to go out and buy this specific part from this specific company for their builds, that would be against the law. I do believe Intel tried to do similar practices many times and got slapped for it repeatedly. But the only way for a company to force a product to be made exactly one specific way is to make it themselves.

Just like Generic Drugs. They just have to follow specific guidelines and rules but not the exact formula. You're ignoring this fact is not going to make it better. There ARE cheap components out there and there are companies that pay a premium for the best parts the manufacturer can deliver. They can even be the exact same parts, just the best of the best picked for the premium company, the rest (repaired-to-work parts and not-quite-perfect but fully otherwise functional parts) go to the budget companies. So they can even have the exact same model/type numbers on them but be truly completely different parts.

Ask anyone in the electronics manufacturing or silicon industry if you don't want to believe me. This is extremely common practice. I know for a fact the top companies like Intel/Nvidia/AMD does it (I used to work for AMD directly in relation to the bin-out process, also called the 'sort' process). And the smaller-part venders have similar quality validation processes.

Heck for that matter, most of your parts are just not-fully-functional other parts. Like say a tri-core chip from AMD. It's really just a failed quad-core where one core failed validation. So they turn it off and mark it as a tri-core. Or a 3.2Ghz part just wasn't quite stable enough for them to comfortably stamp it as aq 3.4Ghz part. But it's the exact same part. All another example of the bin-out process, just in a slightly different form. But even in say the memory (which my current company makes tons of: Spansion), there can be tons of levels of quality output for the same exact chip of memory.

You can see this represented partially in RAM chips even with their CAS speed capabilities. The chips will be manufactured exactly the same but this chip tests better than that one so it will end up on a CAS 6 stick while the poorer one is limited to a CAS 9 use only... Or, to be a bit more accurate; this company buys only our top-shelf chips to use on their gaming sticks while the budget company buys them all and tries to force them higher than we recommend. Exact same part but not necessarily capable of the exact same performance.

So yes, absolutely positively and without question, there are quality components and budget components, even if they have the exact same part number. And if you don't want to take my word on it; ask anyone in the industry. These are standard practices following specifically laid out procedures and standards throughout the world. We even have parts we label as (Well, it's a code name but it means basically) 'fully functional but...' which means they tested just fine but had a questionable process in manufacturing somewhere along the line. These lots never go to the top-tier clients. They pay extra for the guarantee that they get the best of the best...

So don't fool yourself that the budget cards are exactly the same as a quality card just because they are both a 'reference' design. That just means they follow the reference design, not exactly what parts they use or what grade of those parts they purchased to use. Even two cards made with the exact same types of parts can be completely different from the bin-out process and specific part selection.
 
Last edited:

VirtualLarry

No Lifer
Aug 25, 2001
56,447
10,117
126
But the only way for a company to force a product to be made exactly one specific way is to make it themselves.

So don't fool yourself that the budget cards are exactly the same as a quality card just because they are both a 'reference' design. That just means they follow the reference design, not exactly what parts they use or what grade of those parts they purchased to use. Even two cards made with the exact same types of parts can be completely different from the bin-out process and specific part selection.

I thought that was how "reference" cards worked - the actual card mfg was fabbed out to a contract mfg firm, and they produce thousands or millions of them, and then the vendors purchase the cards wholesale, put their stickers on them, put them into a bright shiny box, and sell them?
 

SirGCal

Member
May 11, 2005
122
1
0
www.sirgcal.com
I thought that was how "reference" cards worked - the actual card mfg was fabbed out to a contract mfg firm, and they produce thousands or millions of them, and then the vendors purchase the cards wholesale, put their stickers on them, put them into a bright shiny box, and sell them?

Nope, I tried to explain that in the last post... But the reference is just a schematic of how to build the card. Purchasing the individual components for the card is still up to the manufacturer. Kind of like an engine, the blueprints will say you need this type of block, this size pistons, lifters, valves, body, etc. But which ones exactly you choose to purchase are up to you. Following the same engine schematic to the letter but changing the specific parts can change the resulting engine power and efficency a ton.

Think of the reference as a spec. For example say part of the reference is to have +/- 0.5V on the 12V input for example. One company might just barely make that work while another company might go further to limit it to +/-0.1V, going above and beyond the specifications alone. That's another way they can make reference better. But the biggest part directly is just which parts they buy for the build. For example in the simplest fassion; are they buying the most budget line of 'Cayman' chips from AMD (the actual 69xx series GPU chip) or are they paying more for a higher-grade bin-out part? The pick of the litter, so to speak. Do they spend a bit more to assure they do not get any of the 'fully function but...' parts and only pure strait-up passing parts... And that's just one example. But it goes from the big chips all the way down to the smallest resisters... It might be once cent per hundred parts vs one cent per ninety parts, but it adds up when you're making thousands of cards with hundreds or thousands of individual parts.

I hope that helps make more sense.
 

SirGCal

Member
May 11, 2005
122
1
0
www.sirgcal.com
And it's not just video cards, it's ALL electronics... Hence buying a name, is not always buying a name. Infact in the electronic industry, many of them are simply buying the better parts. Everything from cellphones (though there isn't much direct competition for that example but...) to things you don't even think about (the computer controlling your car's engine for example). Sure some names are just that... but many of the 'slightly more expensive' main brands are the ones you really want to look at; the ones buying the higher-bin-out parts.
 
Status
Not open for further replies.
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |