Known/suspected games to eat more than 1GB video memory at 1920x1200

bhavv · 9 Apr 2011 at 22:47

drifting said:
my min fps is 44% higher than your best min fps, while it's just pushing your vram to the edge.

So your min FPS is 0.44 x 9 + 9 better = 4 FPS more for 13 FPS total?

What do you want for that, a medal or a blue peter badge? The game would still be unplayable with such a low FPS, and I already stated several times that 2 Gb Vram only makes a difference of a few mostly unnoticable FPS over 1 Gb ram, and you simply prove that with your absolutely amazing (to you only) 13 FPS minimum on your 2 gb cards.

Also I dont see your results anywhere, and you never asked me to run the menchmark fully maxed out, nor specified what settings to use:

drifting said:
Post your results of metro2033benchmark.exe then.

I did that at your request, and all you do is whinge and whine. Do you really think that anyone else here is taking you seriously? Because no one seems to be believing anything you say about the issue of Vram.

TBH arguing this point with Raven was more interesting than it is with you, I want Raven back to argue about 2 Gb Vram with, not you.

drifting · 9 Apr 2011 at 22:50

bhavv said:
So your min FPS is 0.44 x 9 + 9 better = 4 FPS more for 13 FPS total?

What do you want for that, a medal or a blue peter badge? The game would still be unplayable with such a low FPS, and I already stated several times that 2 Gb Vram only makes a difference of a few mostly unnoticable FPS over 1 Gb ram, and you simply prove that with your absolutely amazing (to you only) 13 FPS minimum on your 2 gb cards.

This is an unfair comparison, when you have (intentionally) turned off DOF and used 4AF. Can I say that even an HD5750 @ 640x480 with min settings can beat GTX580 @ 2560x1600 with max settings?

I even posted my results but I don't understand why you don't even bother to have a look. You have been claiming that you can run games with max settings, but yet you don't know what exactly does "max settings" mean? :S

If you still can't see my post of my results, I can specify the settings clearly here: DX11, 1920x1200, Very High, MSAA 4X, AF 16X, DOF Enabled, Advanced PhysX Disabled

bhavv · 9 Apr 2011 at 22:54

drifting said:
(intentionally)

No, it really wasnt intentional, I double clicked the benchmark, set the resolution, DX11 and AA, but forgot about the other settings because normally you dont set anything else in other benchmarks.

drifting said:
You have been claiming that you can run games with max settings, but yet you don't know what exactly does "max settings" mean? :S

I go into the options menu and turn everything up to full, and everything works? I'm not sure if I played enough of Metro to get to any slowdown bits, as I said I only played it for about an hour. Actually, 44 inutes played according to steam. I can play all my other games with all the options set to full.

drifting · 9 Apr 2011 at 22:55

bhavv said:
I did that at your request, and all you do is whinge and whine. Do you really think that anyone else here is taking you seriously? Because no one seems to be believing anything you say about the issue of Vram.

TBH arguing this point with Raven was more interesting than it is with you, I want Raven back to argue about 2 Gb Vram with, not you.

No, but the one who knows how expensive DOF is, would laugh at your results, because it doesn't even mean anything here. You could hide from my thread and play without DOF and enjoy your sufficient 1GB.

bhavv said:
No, it really wasnt intentional, I double clicked the benchmark, set the resolution, DX11 and AA, but forgot about the other settings because normally you dont set anything else in other benchmarks.

I go into the options menu and turn everything up to full, and everything works? I'm not sure if I played enough of Metro to get to any slowdown bits, as I said I only played it for about an hour.

OK sorry then, it was my fault to fail to understand that you didn't know about what exactly is called "max" settings for each option.

Let's not start a war in this thread. I have mentioned that the stage "Child" in Chapter 4 can be stressful. You most probably haven't reached that scene yet. Then fine, I ask you to run the less stressful thing, which is the benchmark, with max settings. If you dare not even run the benchmark with appropriate settings, then how would you make your criticisms constructive?

bhavv · 9 Apr 2011 at 22:58

drifting said:
No, but the one who knows how expensive DOF is, would laugh at your results, because it doesn't even mean anything here. You could hide from my thread and play without DOF and enjoy your sufficient 1GB.

Huh? But if 1 Gb Vram isnt even sufficient with DoF disabled and with only 934 Mb Vram being used up, how would 2 Gb Vram be anymore sufficient?

If I had 2 Gb GTX 560s instead of 1 Gb ones, I dont believe the game would run any better as you claim it would.

tom91x91 · 9 Apr 2011 at 22:58

lol my 6950's can run metro full aa dof, dx11 fine never drops below 30, and is usually 50-60. Looks great.

bhavv · 9 Apr 2011 at 23:00

tom91x91 said:
lol my 6950's can run metro full aa dof, dx11 fine never drops below 30, and is usually 50-60. Looks great.

My cards did run the game fine too, I never noticed anything going below 30 fps during the 44 minutes I played, but the benchmark stresses your PC far more than the actual game does.

bhavv · 9 Apr 2011 at 23:03

God damn you complete jokers, I had advanced Physx on you silly ATI derps

Lets see your 6950 results with advanced physx enabled .... oh no, wait a minute ....

And yes, I was playing the actual game with Physx enabled too. Can you do that? Erm, nope

drifting said:
No, but the one who knows how expensive DOF is, would laugh at your results, because it doesn't even mean anything here. You could hide from my thread and play without DOF and enjoy your sufficient 1GB.

I bet that my use of physx was even more expensive.

drifting · 9 Apr 2011 at 23:06

bhavv said:
My cards did run the game fine too, I never noticed anything going below 30 fps during the 44 minutes I played, but the benchmark stresses your PC far more than the actual game does.

So you keep talking here instead of showing us a reasonable benchmark result. Can I believe that you actually have already got the result but dare not publish it?

drifting · 9 Apr 2011 at 23:08

bhavv said:
God damn you complete jokers, I had advanced Physx on you silly ATI derps

Lets see your 6950 results with advanced physx enabled .... oh no, wait a minute ....

And yes, I was playing the actual game with Physx enabled too. Can you do that? Erm, nope

I bet that my use of physx was even more expensive.

If you think PhysX is expensive, you could disable it, and when you disable it, it wouldn't affect your claim about vram usage. This is not a war between the green team and the red team, this is about vram usage.

bhavv · 9 Apr 2011 at 23:14

I disabled Physx and enabled DOF and 16x AA.

The whole benchmark was smooth with no hitching, and the FPS counter only dipped to around 20. Maximum Vram used was 1014.

HOWEVER, after completing the benchmark, the results stated that my minimum FPS was 5. I can absolutely guarantee that I never saw any such lag spikes or slowdown lower than 20 fps during the benchmark, while it was a lot more laggy and slow throughout with Physx enabled.

drifting · 9 Apr 2011 at 23:17

bhavv said:
I disabled Physx and enabled DOF and 16x AA.

The whole benchmark was smooth with no hitching, and the FPS counter only dipped to around 20. Maximum Vram used was 1014.

HOWEVER, after completing the benchmark, the results stated that my minimum FPS was 5. I can absolutely guarantee that I never saw any such lag spikes or slowdown lower than 20 fps during the benchmark, while it was a lot more laggy and slow throughout with Physx enabled.

You still dare not post the results here. As long as you use every setting exactly as I specified, except PhysX (which I leave it up to you whether to enable it or disable it, but for apple-to-apple comparison I strongly recommend you disable it, while saving you some cost), dare you post your result here? Or you need some extra time to *tweak* the results?

Can I guess, ah wait, is it also affecting your average fps obviously as well?

Once you try this benchmark with 560 2GB, you would most probably see vram usage above 1GB. Now you only see 1GB vram usage because you only have 1GB.

I would trust the min fps numbers more than trust your swear about your subjective feelings here.

bhavv · 9 Apr 2011 at 23:21

drifting said:
You still dare not post the results here.

No it just takes a while to copy / paste all the screens.

The benchmark actually ran fine, that min FPS result couldnt ever be seen while it was running, as you can see on the graph, and I didnt notice any limitation with only 1 Gb Vram.

drifting said:
Once you try this benchmark with 560 2GB, you would most probably see vram usage above 1GB. Now you only see 1GB vram usage because you only have 1GB.

Wow, you dont say captain obvious. However, what I have said throughout this thread is that such an increaded Vram usage on a 2 Gb card would NOT improve performance by anymore than a few FPS.

drifting · 9 Apr 2011 at 23:26

bhavv said:
No it just takes a while to copy / paste all the screens.

The benchmark actually ran fine, that min FPS result couldnt ever be seen while it was running, as you can see on the graph, and I didnt notice any limitation with only 1 Gb Vram.

OK fine, it seems you have finally made it, to successfully use the settings I specified. Oh wait, what just happened to your average fps and min fps?

Still insisting that 1GB vram is sufficient?

drifting · 9 Apr 2011 at 23:27

bhavv said:
Wow, you dont say captain obvious. However, what I have said throughout this thread is that such an increaded Vram usage on a 2 Gb card would NOT improve performance by anymore than a few FPS.

Didn't someone just mention that his super overclocked 560 can be faster than 570?

Where is your claim about 560 on par with 6950?

bhavv · 9 Apr 2011 at 23:27

How about you learn to read? How would 2 Gb ram make a significant difference to the min and average FPS?

drifting said:
Didn't someone just mention that his super overclocked 560 can be faster than 570? Where is your claim about 560 on par with 6950?

Already posted right here:

http://www.bit-tech.net/hardware/graphics/2011/01/27/msi-geforce-gtx-560-ti-1gb-review/8

Also, that comment had nothing to do with Vram either.

drifting · 9 Apr 2011 at 23:29

bhavv said:
How about you learn to read? How would 2 Gb ram make a significant difference to the min and average FPS?

Already posted right here:

http://www.bit-tech.net/hardware/graphics/2011/01/27/msi-geforce-gtx-560-ti-1gb-review/8

It would help your super overclocked 560 maintain the same average fps and min fps as a GTX 570. With only 1GB, it's not possible even in this tiny benchmark.

bhavv · 9 Apr 2011 at 23:30

drifting said:
It would help your super overclocked 560 maintain the same average fps and min fps as a GTX 570. With only 1GB, it's not possible even in this tiny benchmark.

Where are the GTX 570 results for this benchmark? How do you know that the performance difference is due to the Vram and not the extra Cuda cores, which would boost performance a lot more, particularly with Physx enabled in a game like Metro.

You cant make a Vram comparison with two different GPUs, isnt that obvious to you?

drifting · 9 Apr 2011 at 23:35

bhavv said:
Where are the GTX 570 results for this benchmark? How do you know that the performance difference is due to the Vram and not the extra Cuda cores?

You cant make a Vram comparison with two different GPUs, isnt that obvious to you?

We already know that 570 is generally faster than 6950. Your 560 is not even faster than my 6950. Either 570 is slower than 6950 (which doesn't seem true, not even having to mention that Metro 2033 is an nVidia's game), or your claim about 560 is on par with 570 is false.

You still don't admit that the tiny 1GB on your 560 is bottlenecking your benchmark results. But others would clearly have got an idea here.

tom91x91 · 9 Apr 2011 at 23:36

is there a download link for the metro benchmark? Cant find it in my metro folder.