NVIDIA GeForce GTX 580 1536 MB

[HTC]

Any chance W1zzard can add to the review what one gets when the "limiter" on the car is off?

I just have a couple of questions:

- Are there any performance gains when not limiting?

- Is it worth it to remove the limiter?

 

[newtekie1]

Any chance W1zzard can add to the review what one gets when the "limiter" on the car is off?

Performance in everything but Furmark would be unchanged since Furmark was the only application that the limitter detects and activates with. All other apps are already not effected by the limitter.

 

[CDdude55]

Performance in everything but Furmark would be unchanged since Furmark was the only application that the limitter detects and activates with. All other apps are already not effected by the limitter.

Hmm, so Unigine Heaven isn't affected by the throttling?

 

[HTC]

Performance in everything but Furmark would be unchanged since Furmark was the only application that the limitter detects and activates with. All other apps are already not effected by the limitter.

Really?

I thought the extra power it uses when the limiter is not in use would bring some benefits in performance.

 

[newtekie1]

Hmm, so Unigine Heaven isn't affected by the throttling?

Really?

I thought the extra power it uses when the limiter is not in use would bring some benefits in performance.

It amazes me sometimes that people can't be bothered to actually read the review they are commenting on...

At this time the limiter is only engaged when the driver detects Furmark / OCCT, it is not enabled during normal gaming.

 

[HTC]

It amazes me sometimes that people can't be bothered to actually read the review they are commenting on...

OK: where in the review, other then the power consumption test's last graph, does it show any difference, if any, with the limiter on VS off?

 

[CDdude55]

It amazes me sometimes that people can't be bothered to actually read the review they are commenting on...

I did read the review, and Unigine isn't normal gaming load, so it's a valid question. Chill out.

 

[newtekie1]

OK: where in the review, other then the power consumption test's last graph, does it show any difference, if any, with the limiter on VS off?

Seriously, it is not that hard to understand. The driver detects the exe for OCCT or Furmark and enabled the limitter. Nothing else is effected, so your question about performance is answered in the review already.

I did read the review, and Unigine isn't normal gaming load, so it's a valid question. Chill out.

No it is not a valid question, because again, it says right in the review that the driver only activates the limitter when it detects OCCT or Furmark. Is Unigine OCCT or Furmark? No. So there is your answer.

 

[Steevo]

Not true or else I could change the name. The purpose is to throttle in what it considers a overpower situation, thus it throttled during overclocking as shown in the review.

 

[newtekie1]

Not true or else I could change the name. The purpose is to throttle in what it considers a overpower situation, thus it throttled during overclocking as shown in the review.

Wanna take a stab at how W1z likely bypassed the limit to get Furmark numbers w/o the limitter?

And no where in the review does it say the limitter was activated during overclocking.

Yes, it is possible that it could be set up to throttle in every overpower situation. However currently, as W1z stated, it only monitors the sensors when OCCT or Furmark is detected running. Those two applications are the only application effected. No ther application is effected by the limit currently.

 

[HTC]

Seriously, it is not that hard to understand. The driver detects the exe for OCCT or Furmark and enabled the limitter. Nothing else is effected, so your question about performance is answered in the review already.

It's amazing how someone can comment on a review without actually bothering to read it ...

Once we reached 97°C the card started to throttle down which forced us to lower clocks to reduce power consumption to get out of the throttling state to maximize performance.

When the card senses it is overloaded by either Furmark or OCCT, the card will reduce clocks to keep power consumption within the board power limit of 300 W

How do you know if, in whatever game, the card reaches the 97º limit and starts throttling down or not without actually testing for it? I dunno, since i don't own the card.

If not W1zzard, perhaps someone who does own such card can test this and report back.

 

[CDdude55]

It's a valid question newtekie.

For the every day gamer the power draw limiter will not have any effect on performance.

As mentioned earlier, the card comes with a current limiter system which reduces clocks and performance in case the card senses it is overloaded.

You're basing what you're saying on the fact that it is driver limited for those two specific benchmarks, but it has also been stated that when overloaded the card will throttle, so if a program like Heaven is strong enough (because it's not regular load) to push the temps up significantly, then performance throttling would be expected. I'm not saying you're wrong and that the drivers don't dictate what's regular load and whats not, but based on a variety of different statements it wasn't 100% clear cut to me. Because as i said, Unigine isn't gaming load, so it's more then a valid question. No need to start PMSing lol, just a joke.

 

[Steevo]

Or a game, or any CUDA program that pushes the card.


I am not saying it is specifically a bad thing, perhaps they plan on cooking some older cards and wanted to protect the new users again?


But my problem is what about the users that want to run this under water, or more extreme situations? Is Nvidia giving them the finger on this round? There has been talk of a "performance boost" on some forums, perhaps they have waiting till AMD drops their card to unleash its extra clocks to "meet the competition"?


I just question the wisdom of tying down a the high end card, targeted at enthusiasts, when they get bad press for it.

 

[newtekie1]

It's amazing how someone can comment on a review without actually bothering to read it ...





How do you know if, in whatever game, the card reaches the 97º limit and starts throttling down or not without actually testing for it? I dunno, since i don't own the card.

If not W1zzard, perhaps someone who does own such card can test this and report back.

The 97°C limit has nothing to do with the current limitter. The 97°C limit has been on all the Fermi cards, AFAIK. Temp is not the same as current.

It's a valid question newtekie.





You're basing what you're saying on the fact that it is driver limited for those two specific benchmarks, but it has also been stated that when overloaded the card will throttle, so if a program like Heaven is strong enough (because it's not regular load) to push the temps up significantly, then performance throttling would be expected. I'm not saying you're wrong and that the drivers don't dictate what's regular load and whats not, but based on a variety of different statements it wasn't 100% clear cut to me. Because as i said, Unigine isn't gaming load, so it's more then a valid question. No need to start PMSing lol, just a joke.

No, I'm basing what I'm saying on the fact that W1z has said that the limitter is only active when those two programs are detected by the driver. In all other cases the driver doesn't care about current and doesn't activate the limitter even if the card goes over 300w.

Again, it can't get more clear cut than "the limiter is only engaged when the driver detects Furmark / OCCT", at no other point will the limitter activate, even if the application/game causes the current to go over 300w.

Or a game, or any CUDA program that pushes the card.


I am not saying it is specifically a bad thing, perhaps they plan on cooking some older cards and wanted to protect the new users again?


But my problem is what about the users that want to run this under water, or more extreme situations? Is Nvidia giving them the finger on this round? There has been talk of a "performance boost" on some forums, perhaps they have waiting till AMD drops their card to unleash its extra clocks to "meet the competition"?


I just question the wisdom of tying down a the high end card, targeted at enthusiasts, when they get bad press for it.

Considering we already have ways around the 300w limit, I don't think this is the case. I think there is a specific reason they only target 2 programs and don't monitor current during any others. Thats not to say they won't expand the list later on though.

 

[HTC]

The 97°C limit has nothing to do with the current limitter. The 97°C limit has been on all the Fermi cards, AFAIK. Temp is not the same as current.

Dunno: you could be right.

No, I'm basing what I'm saying on the fact that W1z has said that the limitter is only active when those two programs are detected by the driver. In all other cases the driver doesn't care about current and doesn't activate the limitter even if the card goes over 300w.

Again, it can't get more clear cut than "the limiter is only engaged when the driver detects Furmark / OCCT", at no other point will the limitter activate, even if the application/game causes the current to go over 300w.

Considering we already have ways around the 300w limit, I don't think this is the case. I think there is a specific reason they only target 2 programs and don't monitor current during any others. Thats not to say they won't expand the list later on though.

Since W1zzard didn't say that (bold underlined), your other points (bold) are wrong.

This is what W1zzard said:

Once we reached 97°C the card started to throttle down which forced us to lower clocks to reduce power consumption to get out of the throttling state to maximize performance.

When the card senses it is overloaded by either Furmark or OCCT, the card will reduce clocks to keep power consumption within the board power limit of 300 W

If you're right, anyone with the card can test this by changing Furmark's EXE file to ... "x.EXE" or something like that: then, he / she will get quite a hike in temps.

Maybe W1zzard can clarify it for us: has the card been limited by those 2 programs specifically? I ask because he said "senses it is overloading" and this can be done by other programs as well: will the limiter kick in then too?

 

[wahdangun]

The 97°C limit has nothing to do with the current limitter. The 97°C limit has been on all the Fermi cards, AFAIK. Temp is not the same as current.



No, I'm basing what I'm saying on the fact that W1z has said that the limitter is only active when those two programs are detected by the driver. In all other cases the driver doesn't care about current and doesn't activate the limitter even if the card goes over 300w.

Again, it can't get more clear cut than "the limiter is only engaged when the driver detects Furmark / OCCT", at no other point will the limitter activate, even if the application/game causes the current to go over 300w.



Considering we already have ways around the 300w limit, I don't think this is the case. I think there is a specific reason they only target 2 programs and don't monitor current during any others. Thats not to say they won't expand the list later on though.

sp if the driver just check furmark.exe then why wizz bother to build GPU-z to overcome the limiter, and every reviewer can just rename the exe and move on

 

[LAN_deRf_HA]

After giving it some thought, I think this limiter thing is perfectly acceptable as long as they include an official option to disable it in the control panel, or put some documentation somewhere telling us it can be done with gpuz. This way they protect the average consumer and stay within the pci-e rating, but give the option to go beyond to the people who actually know what they're doing. Atm though I don't see them doing that, if anything I wonder if they'll block the gpuz bypass in the next driver.

 

[HalfAHertz]

Well only stress testing programs use it, so I don't see a need to disable it.I mean if all you're using the card for is gaming, and you want to over-clock, then why not stress test with a demanding game instead? Sometimes I've had OCs which would have corruptions and artefacts in The Furry Donut™, but would work perfectly in games.

 

[newtekie1]

Dunno: you could be right.


Since W1zzard didn't say that (bold underlined), your other points (bold) are wrong.

Oh Jesus Christ!!! I quoted him saying exact that! Do you have a hard time reading? I'm not quoting him again, you can read my post above to see where he said it, or even better read the review!:shadedshu

This is what W1zzard said:

Once we reached 97°C the card started to throttle down which forced us to lower clocks to reduce power consumption to get out of the throttling state to maximize performance.

Again, why are you taking a statement from the overclocking section about temperature and trying to say it has anything to do with the Over Current Protection? Did you not get that the temperature protection is a totally different thing from the current protection? I believe I already told you this.

You can read pretty much the same statement about GTX480s at well: http://www.techpowerup.com/reviews/MSI/N480GTX_GTX_480_Lightning/31.html

But I guess you will still want to go on about the temp limit like it is the same thing as the current limit, but now I'm sure you'll say the GTX480 must have had it too...right? Because this is like the 3rd time you've went on about the temperature limit like it is related to the current limit.:shadedshu

Again, temperature is not the same as current, they are two different things and hence two different protection systems.

When the card senses it is overloaded by either Furmark or OCCT, the card will reduce clocks to keep power consumption within the board power limit of 300 W

Correct he did say that, and I bolded the important part. And when it is put in context with the fact that he said the OCP is only activated when it detects OCCT and Furmark.

If you're right, anyone with the card can test this by changing Furmark's EXE file to ... "x.EXE" or something like that: then, he / she will get quite a hike in temps.

Maybe W1zzard can clarify it for us: has the card been limited by those 2 programs specifically? I ask because he said "senses it is overloading" and this can be done by other programs as well: will the limiter kick in then too?

Again, I don't know why you can't be bothered to read the review as W1z already said this in it. The driver detects Furmark and OCCT and activates the limitter. It is limitted by those 2 programs specifically because those are the only two programs the driver detects. In all other programs the driver doesn't monitor the overcurrent sensors.

sp if the driver just check furmark.exe then why wizz bother to build GPU-z to overcome the limiter, and every reviewer can just rename the exe and move on

Because when nVidia adds more programs latter it will be easier to just use the GPU-z tool to disable the limitter instead of trying to figure out if a program is affected and what programs are affected.

Plus there are more ways than just the exe name to detect if a certain program is running.

 

[bear jesus]

Sometimes I've had OCs which would have corruptions and artefacts in The Furry Donut™, but would work perfectly in games.

But does that not just mean it is not really 100% stable, kind of like a cpu overclock that is fine in games but something like prime causes bsod's?

I must admit though with my 6870's i have had overclocks that were furmark and atitool stable with no artifacting yet 3dmark06 would cause crashes so it's not like furmark is a definitive stress test.

 

[N3M3515]

From my experience with graphic cards, all of them have a temperature limit, once they reach it they lower the clocks, i've had 5700 ultra, 6800 gs, and when they reached 97 - 100 degrees, the driver throtled the clocks down, so for those thinking it was the OCP from the GTX 580, it's not.
The OCP works only when reading current passes 300W(my guess), but since there are no games that makes it surpass 300W then it only works when the only apps that makes the card surpass 300W: occt and furmark.

There you go, temp is not the same as current OCP reads current: Watts

 

[Steevo]

What happens once a game comes along that uses more than 300W, nvidia just expects users to live with underperformance then? Or do they expect you to upgrade cards?


Would you buy a car with a wood block under the throttle?

 

[N3M3515]

What happens once a game comes along that uses more than 300W, nvidia just expects users to live with underperformance then? Or do they expect you to upgrade cards?


Would you buy a car with a wood block under the throttle?

I think they will take care of it when it happens, driver level.

 

[Benetanegia]

What happens once a game comes along that uses more than 300W, nvidia just expects users to live with underperformance then? Or do they expect you to upgrade cards?


Would you buy a car with a wood block under the throttle?

That will never happen. Never. It's impossible to load all the resources on a GPU at the same time unless you artificially do it with something like Furmark. In actual games, there's a lot of things to be made other than running shader code, like for example texture loading and filtering.

 

[Steevo]

So then they felt the extra money they could spend on more hardware was obviously worth it for the few who would run furmark or occt? Bullshit.

The did it to either mislead the public in power use, or to protect the card from being used to the full in a optimized way.



A car with a wood block under the gas pedal.