Benchmark for Black Shark?

Just thought I'd add something for people without Vista:

The windows 7 beta is running until Aug 9th and now has unlimited keys, all people with multi cores should do a dual boot for it to try BS.

By then there should be true multicore support for all operating systems in BS.

I still had the origninal stress test track, looks like it got removed or accidentally deleted. Here is the one I believe was posted orginally if you missed it.

[ATTACH]23687[/ATTACH]

Thanks!

Still i'd like to know it clear if EN & RU release versions of the game do have some difference with it.

I still had the origninal stress test track, looks like it got removed or accidentally deleted. Here is the one I believe was posted orginally if you missed it.

 

[ATTACH]23687[/ATTACH]

Thanks for posting that. On my machine, Win 7, same settings as my previous 3.6GHz run, I got the following:

2009-01-14 01:42:35 - dcs

Avg: 85.027 - Min: 29 - Max: 201

For some reason view control stopped being controlled properly within the recording so I stopped the benchmark at that point. Not sure what that was about. Is it intentional to have the map screen included during playback? Seems strange, that must be where the 201 fps came from.

[edit] Oh and Maximus, I can't recall where I read it, but it was a source I thought I could trust who stated that the smoke of the flares was tweaked between the Russian and English release such that it didn't impact the frame rate in the English version nearly as much as the Russian release did. That might not be accurate though!

I still had the origninal stress test track, looks like it got removed or accidentally deleted. Here is the one I believe was posted orginally if you missed it.

 

[ATTACH]23687[/ATTACH]

Not sure why but I don't get consistent results when using this track with FRAPS like I do with other .trks... :huh:

:( Fraps .. hummmm

The track have some View controls problems as well here

Does Fraps have a performance impact on the game?

There is a small overhead associated with drawing the framerate on screen. This may vary depending on your system configuration, but should remain relatively minimal. The best way to measure it on your own system is to find a game that allows you to benchmark it and compare the results obtained with and without Fraps loaded.

When you are benchmarking the overlay is automatically disabled to provide the most accurate results.

( Russian version ) I did a quick one.. but i stop just before it get to the bad views. ( all setting at max but water one notch down since i don't like the mirror effect)

Avg: 124.005 - Min: 96 - Max: 164

It goes to the F10 map view for me also. Not sure about the inconsistent results but a few times my HD appeared to be loading something and would give a small hickup in the game.

Also the new 181.20 Nvidia drivers are supposed to support multi-monitor SLI. Has anyone benchmarked or tried this yet? Kind of curious how good it is. Seems like it would render faster with each card rendering to its own monitor.

based on the reportedly poor performance of the GTX 260/280 series cards in FSX I have kept my 8800GTS which does well in most applications I run.

I've been beta testing a new flight sim and wanted to see how the new video cards would handle so I picked one up. I also did a quick test in BlackShark. Here are the results.

Using my previous benchmark settings, but swapping to the GTX 285SC and 185.85 Drivers I see the following performance data

********8800GTS (512) vs GTX 285 SC (1024)********

Windows7 (64) core 0+1 E8400 @ 3.960GHz 2x2Gigs of PC8500 RAM at 1100MHz, 8800GTS (G92) at stock clocks 178.24 drivers vSynch OFF

Avg: 58.823 - Min: 32 - Max: 93

Windows7 (64) core 0+1 E8400 @ 3.960GHz 2x2Gigs of PC8500 RAM at 1100MHz GTX 285SC 675 MHz core, 1548 shader clocks, 2538 memory 185.85 WHQL drivers.

Avg: 51.018 - Min: 22 - Max: 93

Unfortunately it seems that the GTX 285 is much slower at these settings than even a lowly 880GTS 512 in BlackShark. I lost 7.805 fps average, and worse yet, my minimum fps dropped by 10! The New drivers are supposedly very fast in most modern applications, however they may partially to blame here, but saw similar loses in FSX when running the "faster" card.

OK, so it's slower at those settings, but how does the 285 handle when Driver level settings are maxxed out?

Windows7 (64) core 0+1 E8400 @ 3.960GHz 2x2Gigs of PC8500 RAM at 1100MHz 16xAA 16xAF, High Quality, no optimizations, GTX 285SC 675 MHz core, 1548 shader clocks, 2538 memory 185.85 WHQL drivers.

Avg: 50.796 - Min: 22 - Max: 85

If there is any good news here, it is that turning cranking up the driver level quality to the maximum, and setting AF and AA at 16X has next to no impact on performance. Were it not for the large loss of performance when running the GTX 285 at the lower settings I would possibly conclude that Black Shark is simply CPU limited at 3.96GHz and the video card doesn't matter, but it does as the old 8800 shines at 8XAF/2XAA!

Radeon HD4850 & GTX260-216 comparison in DCS.

Vista, 1920x1080, default driver settings (no AA, AF, VSync), "medium" game settings, i7, affinity set to 2 cores.

At these settings, these video cards show the same picture in the game, no visual differences except chroma/contrast settings.

First, there's a "heavy" track having many units fighting each other, view from the cockpit.

This track is CPU-bound. It shows linear FPS increase, when CPU frequency increases (both frequency and FPS grow at about 30%). At 3.6GHz the game is still CPU-hungry, and the difference between 2 video cards is negligible.

The second track is "stress_test_views". I was expecting to see the same picture here.

Surprisely, GTX showed better at some scenes.

At the start (Ka-50 view), there is no difference.

But later, when the view changes to Mi-28 and then to Mi-24, FPS get better.

It's even much better at "F10" view, though this improvement is not very useful.

The third track is from the demo package, it should be called a "stronghold" or something in the English-language version of the game.

There are a few units, not so many, and here we see the steady 30-70% FPS increase.

So, if the CPU is enough, you can expect a nice improvement in DCS with a more powerful video.

How about some results from people with AMD CPUs.