Frames Time (ms) Min Max Avg

6909 300000 20 27 23.03

 

GTX580 @ 850MHZ

 

Running FSX default 1024 Clouds

 

Thanks Ben :hi:

Frames Time (ms) Min Max Avg

6909 300000 20 27 23.03

 

GTX580 @ 850MHZ

 

Running FSX default 1024 Clouds

Where these tests done using the FSXmark11 settings?

 

Where these tests done using the FSXmark11 settings?

 

No. I made sure to carefully follow the instructions. I am using FSX Default clouds @ 1024. Thats why I scored so well...

No. I made sure to carefully follow the instructions. I am using FSX Default clouds @ 1024. Thats why I scored so well...

 

Thanks for the reply Ben. Would you mind doing the test using FSXmark11 settings (apart from screen resolution) as well?

 

So it's time to recap here, and define a set of FSX and Inspector settings for the test:

These "updated" settings with lowered sliders as posted in the picture above do not work as intended. I believe you defined these settings when you were trying to make sense of Hasses good result with the GTX680.

 

But using the original FSXmark11 settings apart from screen resolution has produced these data points so far from 4 different GPUs in 3 different series used in 3 different computers:

 

I can clearly see a pattern B)

These "updated" settings with lowered sliders as posted in the picture above do not work as intended. I believe you defined these settings when you were trying to make sense of Hasses good result with the GTX680.

 

But using the original FSXmark11 settings apart from screen resolution has produced these data points so far from 4 different GPUs in 3 different series used in 3 different computers:

 

well, I never used FSXmark11 settings there, my first runs where done using my everyday settings including water at Low 2.x, Scenery maxed out, very dense AG and 75% MyTraffic airliners + 25% MyTraffic GA plus 4096 REXE clouds

 

Looks like more factors come into play than I anticipated, but seeing how the 680 - 670 do compared to the variety of Fermis in the sample, I tend to agree it seems to be all about shader umph

well, I never used FSXmark11 settings there, my first runs where done using my everyday settings including water at Low 2.x, Scenery maxed out, very dense AG and 75% MyTraffic airliners + 25% MyTraffic GA plus 4096 REXE clouds

 

Looks like more factors come into play than I anticipated, but seeing how the 680 - 670 do compared to the variety of Fermis in the sample, I tend to agree it seems to be all about shader umph

 

I see. Westman said he used FSXmark11 settings. But I'll ask him to confirm. Never the less, even your settings with a lot of traffic, water Low 2.0 and so on seems to be fully shader limited. Dropping the scenery sliders seems to take us away from beeing shader limited but I'd say most people would be playing with high sliders.

 

EVGA 670 OC 4gb 1186-1166 6600

Frames Time (ms) Min Max Avg

8465 300000 25 33 28.217

Can you remember the core speed on this test? 1186-1166 doesn't really make sense.

 

What are the Factory default speeds for both your GTX670 and your GTX680?

 

Did you use FSXmark11 settings apart for screen resolution on these as well?

 

Cheers

Lars

Firs run with the 680 was FSMark 11 settings.

After that all runs after Dazz Settings.

 

Core Speed 1086- 1166 boost

 

Hasse

Firs run with the 680 was FSMark 11 settings.

After that all runs after Dazz Settings.

 

Core Speed 1086- 1166 boost

 

Hasse

Haha, ok. Thanks. What's the factory default on both of your cards? Thanx for all your testing Hasse

I'm going to throw a wrench at this and will probably do this test this weekend with a GTX280 1GB with my 3770K running at 5Ghz. LOL That should be interesting.

 

- Grammar *censored*, this was sent from my phone.

I'm going to throw a wrench at this and will probably do this test this weekend with a GTX280 1GB with my 3770K running at 5Ghz. LOL That should be interesting.

 

- Grammar *censored*, this was sent from my phone.

Please do. The more results the better. I suspect your card will struggle a bit.

Please do. The more results the better. I suspect your card will struggle a bit.

 

LOL

 

Intel i7-3770K @ 5.0Ghz HT off

2x4GB 2400 (10-12-12-31-1T)

EVGA GTX280 1GB

Samsung 830 SSD 128GB

 

1627, 300000, 4, 6, 5.423

Now with the GTX680 4G installed:

 

Intel i7-3770K @ 5.0Ghz HT off

2x4GB 2400 (10-12-12-31-1T)

EVGA GTX680 FTW LE 4GB

Samsung 830 SSD 128GB

 

Frames, Time (ms), Min, Max, Avg

14030, 300000, 36, 58, 46.767

 

1627, 300000, 4, 6, 5.423

Frames, Time (ms), Min, Max, Avg

14030, 300000, 36, 58, 46.767

Wow, what a difference. Who said FSX isn't limited by the GPU? ^_^

Your results are a lot higher than Westmans also using 4GB GTX680 on highly clocked 3770K with fast RAM. I wonder why...?

 

Lets hope we get some more results comming in from others as well.

 

So far the Keppler cards with a lot of shaders perform a lot better than the Fermi cards.

Wow, what a difference. Who said FSX isn't limited by the GPU? ^_^

Your results are a lot higher than Westmans also using 4GB GTX680 on highly clocked 3770K with fast RAM. I wonder why...?

 

Lets hope we get some more results comming in from others as well.

 

So far the Keppler cards with a lot of shaders perform a lot better than the Fermi cards.

 

Oh yea, no doubt that moving to a Keppler from something like a GTX280 will be a massive improvement. I will be posting the GTX680 results very soon. ^_^

1627, 300000, 4, 6, 5.423

14030, 300000, 36, 58, 46.767

 

 

Holy cow!! :shok: What an improvement! This is solid proof that FSX requires a powerful GPU if you want to do any serious AA.