Furmark+IntelBurnTest (simultaneously) fail (always furmark crash)

[BenchAndGames]

@EarthDog
I understand but I will return it cuz I dont want to start from beggining with troubles. "I mean that first freeze/TDR on idle" for only 20€ more I can actually buy the Ventus OC edition, and hope that one will run good without any problems.
It may sound like I'm a little exaggerating, but I remember few years ago when I got GTX 970 I ahd to replace it also for a new unit cuz I had similars issues, was starting same as this one, but with that one I remember was almost everyday ingames freeze, screen totally blank or green, sound buggy, so I replace that one and the new unit what they send me, was totally fine, 0 problems


@londiste
No, I was able to run it for like 2 hours in solo mode and than I just manually stopped cuz I didnt see any problems.

 

[Shambles1980]

the Only time i use IBT. is when water cooling "custom loop" and overclocking to see what my max fan speeds need to be to keep the system under check for prolonged periods of heavy usage.
I never use IBT when i have a air cooler.
As for IBT + fur mark. i just wouldnt do it. OCCT psu test is enough stress for a system to check for stability over all components.
its a pretty similar setup too. but its automated and runs the tests in an alternating pattern to achieve the results needed for a stability test..

for reall world gaming tests well play the game, or failing that i like firestrike combined test.

 

[EarthDog]

@EarthDog
I understand but I will return it cuz I dont want to start from beggining with troubles. "I mean that first freeze/TDR

@EarthDog
I understand but I will return it cuz I dont want to start from beggining with troubles. "I mean that first freeze/TDR on idle" for only 20€ more I can actually buy the Ventus OC edition, and hope that one will run good without any problems.
It may sound like I'm a little exaggerating, but I remember few years ago when I got GTX 970 I ahd to replace it also for a new unit cuz I had similars issues, was starting same as this one, but with that one I remember was almost everyday ingames freeze, screen totally blank or green, sound buggy, so I replace that one and the new unit what they send me, was totally fine, ....

So, you're guessing it's the GPU based on what a 970 did years ago... interesting.

 

[Regeneration]

Furmark + Linpack is a good way to stress test Northbridge, PSU, and entire bus (FSB/QPI/UPI/HT).

The error is probably related to instability of one of those things.

 

[EarthDog]

Furmark + Linpack is a good way to stress test Northbridge, PSU, and entire bus (FSB/QPI/UPI/HT).

The error is probably related to instability of one of those things.

Furnark is a power virus and should not be run... nvidia and AMD have said this!! It doesnt even test running clocks and voltage and immediately throttles any relatively modern gpus!

Stop saying it is ok...

 

[londiste]

Both Nvidia and AMD have a bone to pick with Furmark because it used to show the maximum power consumption of their cards before they could get their power management in order.

It is perfectly OK to run.
Running clocks are irrelevant. Furmark is to test power consumption and temperatures.

 

[Regeneration]

Nvidia and AMD said it 10 years ago, but FurMark improved ever since. Mining, GPGPU can be stressful like FurMark.

There is nothing wrong with rendering some furry object and textures. If the card can't handle it, something is wrong with its cooler.

 

[EarthDog]

Both Nvidia and AMD have a bone to pick with Furmark because it used to show the maximum power consumption of their cards before they could get their power management in order.

It is perfectly OK to run.
Running clocks are irrelevant. Furmark is to test power consumption and temperatures.

no.

It isnt because they have a bone to pick. They both say it can damage the card and not to run it.

Nvidia and AMD said it 10 years ago,

It's in nvidia reviewers guide for RTX and generations before as well...

 

[londiste]

no.
It isnt because they have a bone to pick. They both say it can damage the card and not to run it.
It's in nvidia reviewers guide for RTX and generations before as well...

Please explain why Furmark is not OK?

 

[Regeneration]

Lousy cooling can damage the card. Not rendering some graphical objects.

 

[EarthDog]

Please explain why Furmark is not OK?

I have already...

But let me get to a desktop (45 mins or so) and I will quote nvidia from the reviewers guide and an FAQ.

 

[Regeneration]

If Nvidia tells you to jump off the roof, will you do it?



Yes, please explain to me how some rotating MSI graphical object can damage the card.

My GTX 970 runs it at TDP + 81 percent. Temps are normal.

 

[EarthDog]

My GTX 970 runs it at TDP + 81 percent. Temps are normal.

So... you don't think your clearly modified TDP on the card has anything to do with that???

ALso, FTR, I haven't mentioned anything about temperatures.



Sorry, it wasn't an FAQ, but from an NVIDIA forums and CSR...

Furmark is an application designed to stress the GPU by maximizing power draw well beyond any real world application or game. In some cases, this could lead to slowdown of the graphics card due to hitting over-temperature or over-current protection mechanisms. These protection mechanisms are designed to ensure the safe operation of the graphics card. Using Furmark or other applications to disable these protection mechanisms can result in permanent damage to the graphics card and void the manufacturer's warranty.

https://forums.geforce.com/default/...r-stress-tests-with-geforce-graphics-cards/1/


Note, I did not add the bold, it is like this in the guide...

For GeForce RTX 2080, we also recommend reviewers test GPU power consumption with actual
games and applications, rather than tools like Furmark.

Applications like Furmark and OCCT—which are often called “power viruses”—will stress the
graphics card beyond normal levels and could permanently damage the card.

You will find that testing power consumption with actual games is much less demanding, and more
representative of the power consumption an end user is likely to see under real conditions.

For certain stress tests such as Furmark and OCCT, if power thresholds are exceeded by the
application, the GPU will go into a reduced performance mode in order to bring power down under
the limits. Like our thermal slowdown mechanisms, this feature is intended to protect the graphics
card and PC from issues caused by too high of power draw.

So again... this is not the first guide that suggests not to run it because it could permantetly damage the card. Yes, there are limits in place on the card, however, how is it a realistic scenario to have your card throttle back clocks and voltage to levels, in my experience recently, hundreds of mhz below the running clock?

Furmark should not be used, period. It can damage cards and is a wholly unrealistic load... even when the car throttles itself back so low to fit in a power envelope. NOT a good test of anything.

Tell me why I should run it again when we are told not to? Do either of you have more anecdotes to add or is straight from the horse's mouth now sufficient?

 

[Regeneration]

This post was from 2011. Some apps and games like The Witcher 3, Fire Strike, can push the GPU to FurMark's level.

https://www.tomshardware.com/reviews/how-to-stress-test-graphics-cards,5449-8.html

Vendors recommend to avoid overclocking,and we still do it.

 

[EarthDog]

Being from 2011 just shows how long they have been saying it... I mean, I just quoted it from the RTX guide...

Vendors recommend to avoid overclocking,and we still do it.

That doesn't mean its right.

Fire Strike does NOT reach FM levels last I ran it... At least, it holds its clocks.

I am running FM now against my 2080 and instantly, my clocks are below the minimum rated boost and still banging off the power limit. They are coming close to base clocks.

 

[Regeneration]

Power:
FurMark 102.2w
Fire Strike 98.8w

Temps:
FurMark 64c
Fire Strike 64c

3.4w is not a big deal.

 

[EarthDog]

You see the difference there, right?


3DM FS runs at clocks and voltages it would normally run at in game. During Furmark, the clocks and voltage are beat down to WELL below running speeds. How can you call that 'like'???

 

[bug]

well,then play those games for heaven's sake

Stupid advice. What's next, "get a real girlfriend"?

 

[rtwjunkie]

Stupid advice. What's next, "get a real girlfriend"?

It’s actually great advice. I gave it earlier too. The best test of your PC’s ability to safely play games is to play those games.

 

[londiste]

So again... this is not the first guide that suggests not to run it because it could permantetly damage the card. Yes, there are limits in place on the card, however, how is it a realistic scenario to have your card throttle back clocks and voltage to levels, in my experience recently, hundreds of mhz below the running clock?

Furmark should not be used, period. It can damage cards and is a wholly unrealistic load... even when the car throttles itself back so low to fit in a power envelope. NOT a good test of anything.

Tell me why I should run it again when we are told not to? Do either of you have more anecdotes to add or is straight from the horse's mouth now sufficient?

Furmark by your own description is an EXCELLECT test of power usage, management and temperatures/cooling.

 

[EarthDog]

It’s actually great advice. I gave it earlier too. The best test of your PC’s ability to safely play games is to play those games.

This is why I don't put too much stock into stress testing CPUs... I give a good few hours in AIDA64 as well as a couple hours of realbench and for MY uses (obviously YMMV) it works out fine for me.

Furmark by your own description is an EXCELLECT test of power usage, management and temperatures/cooling.

How so when it isn't testing the clocks and voltage you run at but is artificially limited by banging HARD off the power limit? Running FS will hit hte limit, but so will any game and the card 'settles' on a boost clock and isn't downclocked BELOW base boost and approaching base clocks (which the card NEVER runs at).

 

[londiste]

How so when it isn't testing the clocks and voltage you run at but is artificially limited by banging HARD off the power limit? Running FS will hit hte limit, but so will any game and the card 'settles' on a boost clock and isn't downclocked BELOW base boost and approaching base clocks (which the card NEVER runs at).

It is a stress test. Pretty much for testing the worst case scenario.
How do you know the clocks, voltage and load you will be running? With games I see this being very-VERY variable.

Besides, what makes one test worse than the other? Tomshardware found Furmark to fall close enough to Witcher 3 which does not sound like an unreasonable thing:
https://www.tomshardware.com/reviews/how-to-stress-test-graphics-cards,5449.html

Most of the games I play have been banging on Power Limit for years, sometimes Voltage Limit as well.

 

[EarthDog]

Again I ask... if you are not testing the same clocks and voltages you are running at (in general) then how is it a worthwhile test? Tell how testing at 1650 MHz .8xxV versus 1950 Mhz 1.05V is a worthwhile stress test?

All cards tickled the power limit in games, but the clocks don't plummet out of the gate to below base boost and approach base clocks.

 

[londiste]

It is not for testing clocks. It is a stress test to test power consumption and temperatures.

 

[qubit]

Runnig Intel Burn Test + Furmark (both last version) simultaneously

That's interesting, it should technically work, but it puts a lot of stress on the system, so it may not be robust enough to take it.

As the others are saying, it could damage your hardware and a bit pointless, so I don't recommend continuing with this.