furmark vs heaven benchmark

[ice856]

hello guys.

 

im just curious, what do you use furmark for?? is it the same as heaven?

i heard that it used to test your overclock setting. but some user also suggest heaven for overclock test.

 

because the interface only shows burn in test and wow it realy make my gpu heat up.

[Si3Rra_7]

I use furmark usually for worst case scenario temp limit

[QueenDemetria]

hello guys.

 

im just curious, what do you use furmark for?? is it the same as heaven?

i heard that it used to test your overclock setting. but some user also suggest heaven for overclock test.

 

because the interface only shows burn in test and wow it realy make my gpu heat up.

I use Furmark for heating my room. Actually, I find Furmark does a better job at heating up my card(which matters to me, since I care about the the thermals under every type of load) and for testing stability. I've had Heaven run perfectly fine with Furmark artifact, so I can't say that Heaven is the best tool for testing OC's.

[SeanAngelo]

furmark is a nononono, i heard furmark has the ability to literally kill your gpu as it exceeds the limit that the card can take.

 

tried furmark once with OC and it reached 82 degrees lololololol

[Najuno]

Furmark is meh... go and use 3D Mark Firestrike, Unigine Heaven 4.0 or Unigine Valley 1.0. These synthetic benchmarks are way better and reflect the performance of a GPU much more precise than other benchmarks. Further there are threads in our GPU sub forum here, where people submitted many results. So it is easy to compare .

[KamiKatze]

Furmark only for quick tests to see the temps or if the OC crashes right away. No longer than 3 minutes. What would i want to run it for? It's not a benchmark.

 

Valley and Heaven do the overclock validation and stability testing.

[Faceman]

hello guys.

 

im just curious, what do you use furmark for?? is it the same as heaven?

i heard that it used to test your overclock setting. but some user also suggest heaven for overclock test.

 

because the interface only shows burn in test and wow it realy make my gpu heat up.

DO NOT use Furmark.  It is rare, but it has been known to brick perfectly good and brand new cards.  Happened to myself.  It is an entirely unrealistic scenario, no other application will ever put as much stress on the GPU as Furmark and it is a dangerous test.  If you want to benchmark your GPU, use Unigine Heaven or other built in game benchmark utilities.

[Dabombinable]

DO NOT use Furmark.  It is rare, but it has been known to brick perfectly good and brand new cards.  Happened to myself.  It is an entirely unrealistic scenario, no other application will ever put as much stress on the GPU as Furmark and it is a dangerous test.  If you want to benchmark your GPU, use Unigine Heaven or other built in game benchmark utilities.

Would it cause a GTX 970 G1 Gaming to degrade after 1 week at +150MHz, +12mV and +450MHz memory? Because to test for instabilities I used MSI Kombuster's version (as in the furry doughnut) for 1 hour to test each overclock (fan set to 100%).

[Najuno]

DO NOT use Furmark.  It is rare, but it has been known to brick perfectly good and brand new cards.  Happened to myself.  It is an entirely unrealistic scenario, no other application will ever put as much stress on the GPU as Furmark and it is a dangerous test.  If you want to benchmark your GPU, use Unigine Heaven or other built in game benchmark utilities.

Furmark is IMHO crap, you would need very good VRM and watercooling to stress test the GPU without risk. And completely not necessary when you have stability stress test like you pointed out with Heaven or Valley. If you are just on air I´d go so far and say don´t crank up the resolution in Heaven a lot because it is likley that your GPU runs into thermal throttling. Heaven is very demanding.

[Faceman]

Furmark is IMHO crap, you would need very good VRM and watercooling to stress test the GPU without risk. And completely not necessary when you have stability stress test like you pointed out with Heaven or Valley. If you are just on air I´d go so far and say don´t crank up the resolution in Heaven a lot because it is likley that your GPU runs into thermal throttling. Heaven is very demanding.

Exactly, Ultra everything on Heaven for 15-20min is an excellent stability and heat test.  It is much more comparable to an in-game scenario rather than the pedal to the metal that Furmark is.

 

This next piece if not just for you, but for everyone because I think this is the best analogy I have ever heard that describes what Furmark does.

 

Analogy Time:

 

    When you are gaming, changes are going on in the game, your GPU is basically using "adaptive mode" similar to your CPU.  Depending on the action, or lack thereof that is going on in game, the GPU adapts to that scenario.  It isn't going to use excessive amounts of power on a loading screen, or a cut scene, or something non-demanding.  When something demanding does happen, then the GPU will ramp up to meet that demand.

   Think of a Nascar Race, imagine the car as your GPU. When you are racing,  your car is always adapting to the situation around you.  You speed up, you slow down, you turn, you pit stop, things are going on that your car has to adapt to similar to your GPU when you game. 

 

    Now, what furmark does is not like a game, or a Nascar race.  Furmark is like a drag race.  A pedal to the metal, push the car/GPU as hard as it can possibly go in a straight line, no turning, no stopping, just full on, all out go, go, GO! In a realistic scenario of GPU use, you would never, ever do this.  It is entirely unrealistic.  I understand why reviewers use it, to get a worst case scenario idea, but it is so far outside the real of realism, that it should not be considered. If your cooling cannot keep up with this excessive heat and load applied to the GPU, or there was a bad solder job done at the factory, or any number of things related to power delivery, that is when Furmark has the potential to break or degrade your GPU.

 

For most people, Furmark doesn't do anything to their GPU.  But for a small percentage of people, Furmark has broken their card, including myself.  A quick google and you will see that this is an issue.  Not a big one, but large enough that it needs to be talked about. 

 

DO NOT USE FURMARK.

[Faceman]

Would it cause a GTX 970 G1 Gaming to degrade after 1 week at +150MHz, +12mV and +450MHz memory? Because to test for instabilities I used MSI Kombuster's version (as in the furry doughnut) for 1 hour to test each overclock (fan set to 100%).

I have no clue, I just wouldn't ever do it again.

[Najuno]

Exactly, Ultra everything on Heaven for 15-20min is an excellent stability and heat test.  It is much more comparable to an in-game scenario rather than the pedal to the metal that Furmark is.

 

This next piece if not just for you, but for everyone because I think this is the best analogy I have ever heard that describes what Furmark does.

 

Analogy Time:

 

    When you are gaming, changes are going on in the game, your GPU is basically using "adaptive mode" similar to your CPU.  Depending on the action, or lack thereof that is going on in game, the GPU adapts to that scenario.  It isn't going to use excessive amounts of power on a loading screen, or a cut scene, or something non-demanding.  When something demanding does happen, then the GPU will ramp up to meet that demand.

   Think of a Nascar Race, imagine the car as your GPU. When you are racing,  your car is always adapting to the situation around you.  You speed up, you slow down, you turn, you pit stop, things are going on that your car has to adapt to similar to your GPU when you game. 

 

    Now, what furmark does is not like a game, or a Nascar race.  Furmark is like a drag race.  A pedal to the metal, push the car/GPU as hard as it can possibly go in a straight line, no turning, no stopping, just full on, all out go, go, GO! In a realistic scenario of GPU use, you would never, ever do this.  It is entirely unrealistic.  I understand why reviewers use it, to get a worst case scenario idea, but it is so far outside the real of realism, that it should not be considered. If your cooling cannot keep up with this excessive heat and load applied to the GPU, or there was a bad solder job done at the factory, or any number of things related to power delivery, that is when Furmark has the potential to break or degrade your GPU.

 

For most people, Furmark doesn't do anything to their GPU.  But for a small percentage of people, Furmark has broken their card, including myself.  A quick google and you will see that this is an issue.  Not a big one, but large enough that it needs to be talked about. 

 

DO NOT USE FURMARK.

THIS! Thumbs up .

[ice856]

Exactly, Ultra everything on Heaven for 15-20min is an excellent stability and heat test. It is much more comparable to an in-game scenario rather than the pedal to the metal that Furmark is.

This next piece if not just for you, but for everyone because I think this is the best analogy I have ever heard that describes what Furmark does.

Analogy Time:

When you are gaming, changes are going on in the game, your GPU is basically using "adaptive mode" similar to your CPU. Depending on the action, or lack thereof that is going on in game, the GPU adapts to that scenario. It isn't going to use excessive amounts of power on a loading screen, or a cut scene, or something non-demanding. When something demanding does happen, then the GPU will ramp up to meet that demand.

Think of a Nascar Race, imagine the car as your GPU. When you are racing, your car is always adapting to the situation around you. You speed up, you slow down, you turn, you pit stop, things are going on that your car has to adapt to similar to your GPU when you game.

Now, what furmark does is not like a game, or a Nascar race. Furmark is like a drag race. A pedal to the metal, push the car/GPU as hard as it can possibly go in a straight line, no turning, no stopping, just full on, all out go, go, GO! In a realistic scenario of GPU use, you would never, ever do this. It is entirely unrealistic. I understand why reviewers use it, to get a worst case scenario idea, but it is so far outside the real of realism, that it should not be considered. If your cooling cannot keep up with this excessive heat and load applied to the GPU, or there was a bad solder job done at the factory, or any number of things related to power delivery, that is when Furmark has the potential to break or degrade your GPU.

For most people, Furmark doesn't do anything to their GPU. But for a small percentage of people, Furmark has broken their card, including myself. A quick google and you will see that this is an issue. Not a big one, but large enough that it needs to be talked about.

DO NOT USE FURMARK.

wow so just stick with heaven then.

[Dabombinable]

wow so just stick with heaven then.

By the sounds of it yes.

[Faceman]

wow so just stick with heaven then.

Yup

[ice856]

but how can furmark draw so much power by showing abstract sauron eyes thing??

[Face2Face]

Furmark is a power virus... delete immediately...

[NeverNude]

Valley benchmark, cause I dig the music.