Total War SHOGUN 2 DX11 with built-in benchmark: even more demanding than Metro 2033!

[sniper_sung]

OK then, since no one with a 1GB card bothers to do such dirty work, I do it myself. I rob the 5850 1GB from my gf's computer and install it into mine.

When I tried to run the official benchmark at DX11 all max 1200p 8AA, I got the following error message:



By applying the trick I mentioned above, I could finish the official benchmark without a problem.

1920x1200 8AA:


1920x1080 8AA:


And the plot of the log files:


During the benchmark I observed severe lag spikes, and such lag spikes are also shown in the plot. Personally I would not call such play experience *acceptable*, especially during multiplayer sessions. Anyone who still insists on 1GB cards shall make his own judgement.

As a side note, some review sites got the fps numbers *in game*. I also tried it myself with Fraps and got a much higher fps, because there are blank margins which significantly reduce the resolution.

[cold2010]

@
sniper_sung

seems that you need to know what the actual play and what benchmark

I do something good for yourself and stay away from the use of benchmark, many of which gives a massive lie, Here are some examples

wow it seems amd's gaming evolved really working well in benchmark.. but not enough to hold the crown in game

By applying the trick I mentioned above, I could finish the official benchmark without a problem.

Test and fool the people that they need only 2 GB to run ,

AMD promotes the NEW cards with 2 GB

[zalbard]

Test and fool the people that they need only 2 GB to run ,

AMD promotes the NEW cards with 2 GB

You should look at the performance graph and lag spike notes in the post above first...
Not to mention that this is nowhere near being a high-end resolution nowadays.

[sniper_sung]

@
sniper_sung

seems that you need to know what the actual play and what benchmark

I do something good for yourself and stay away from the use of benchmark, many of which gives a massive lie, Here are some examples




Test and fool the people that they need only 2 GB to run ,

AMD promotes the NEW cards with 2 GB

No thanks, I trust my own tests rather than some random review article which may have fingers in the marketing pots. I have already made it clear why the tests "in game" from the review posted does not mean anything - if you manually start the Battle of Sekigahara (1600) and record the fps of the opening speech with Fraps in background, the resolution is significantly reduced because of margins on screen. My 5850 1GB can even get higher fps than the GTX 560 Ti 1GB did in the review.

I have already said that the benchmark might be how AMD belittles nVidia's tiny sizes of vram on GTX570/470/560/460 because they've chosen a subtle number just right above 1280MB which I do not believe to be a coincidence. I'm tired of such tricks and I'd get a 2GB card and be done with it rather than spending my time on hacks.

[sniper_sung]

@cold2010

Also a gift for you: 5850 1GB running metro2033benchmark.exe before I removed it from my rig

[cold2010]

You should look at the performance graph and lag spike notes in the post above first...
Not to mention that this is nowhere near being a high-end resolution nowadays.

Talking about prevention work built-in benchmark with AA8 and there is great difference between the two

No thanks, I trust my own tests rather than some random review article which may have fingers in the marketing pots. I have already made it clear why the tests "in game" from the review posted does not mean anything - if you manually start the Battle of Sekigahara (1600) and record the fps of the opening speech with Fraps in background, the resolution is significantly reduced because of margins on screen. My 5850 1GB can even get higher fps than the GTX 560 Ti 1GB did in the review

That's your business but facts remain facts

[Olivon]

Well, well ,well ... Sniper_Sung talking shi* and doing usual propaganda ?

What a surprise ...

[sniper_sung]

Well, well ,well ... Sniper_Sung talking shi* and doing usual propaganda ?

What a surprise ...

Don't feel ashamed for your 1280MB vram being turned down by the benchmark, because 1GB vram is doing worse

[cold2010]

@cold2010

Also a gift for you: 5850 1GB running metro2033benchmark.exe before I removed it from my rig

Super Mario Will support DX11

Recommended System Requirements

OS: Windows 7-Windows Vista
Processor: Intel Core 2 Quad 2.4Ghz, AMD Phenom X3 2.1Ghz
Memory: 2 GB (Windows 7) 2.5 GB (Windows Vista)
18 GB Free Hard Drive Space
Video Card: 15 Gb

Some pictures from within the game

[sniper_sung]

Toki Tori will beat Super Mario in DX9

PC System Requirements
OS: Windows XP/Vista/7
Processor: Intel Pentium 4 / AMD Athlon XP 1.5 Ghz or higher
Memory: 1 GB RAM (Vista) 512mb (XP) or more
Graphics: DirectX 9 compatible, 1280MB VRAM
DirectX®: DirectX 9.0c
Hard Drive: 200 MB of free space
Controller: Microsoft Xbox 360 controller or compatible

[Olivon]

Don't feel ashamed for your 1280MB vram being turned down by the benchmark, because 1GB vram is doing worse

All I see is that your 2GB card are not doing better than my 1280MB card on 1080p.

Your test means nothing because you're completly biaised mrs VRAM and no need to do copy/paste on several sites, this won't make your card better ...

http://www.techpowerup.com/forums/sh...d.php?t=145496

http://we.pcinlife.com/thread-1668615-1-1.html

http://www.chiphell.com/thread-195487-1-1.html

That's why I'm talkin' propaganda and that's the same for your ridiculous 1080p ram usage thread ...

[sniper_sung]

All I see is that your 2GB card are not doing better than my 1280MB card on 1080p.

Your test means nothing because you're completly biaised mrs VRAM and no need to do copy/paste on several sites, this won't make your card better ...

http://www.techpowerup.com/forums/sh...d.php?t=145496

http://we.pcinlife.com/thread-1668615-1-1.html

http://www.chiphell.com/thread-195487-1-1.html

That's why I'm talkin' propaganda and that's the same for your ridiculous 1080p ram usage thread ...

I agree it's "biased", since I only use max settings which is not appropriate for single GPU setup. Originally I thought 1280MB was not enough for Shogun 2 DX11 max because the official benchmark says insufficient vram. However after spamming several forums and asking for test results, it proves me wrong. The current conclusion is that 1280MB is still enough for 1200p 8AA benchmark (no obvious evidence of vram shortage) if you force to override vram value manually, while 1024MB is likely to be insufficient for 1200p 8AA benchmark. However I'm sure 1280MB is close to the edge, and even 1.5G can be exceeded in worst case.

[sniper_sung]

Talking about prevention work built-in benchmark with AA8 and there is great difference between the two

That's your business but facts remain facts

Facts: by default the game automatically downgrades graphics settings if the settings selected cause the graphics card's VRAM to be exceeded.

http://forums.totalwar.com/showthrea...LY-using-Ultra

[sniper_sung]

Thus having little vram is indeed an "advantage" here, because the game would downgrade the graphics settings in stealth, while providing better framerates than the 2GB counterparts running at Ultra.

[NapalmV5]

1280mb is still enough for 1200p 8aa ? hahaha

so why do the game use up to 1500mb out of gtx580 1536mb @ 1080p 8aa ??

It's not that I don't trust your screenshots, but it's hard for me to use my cards to reproduce the scenes you used, that's why a benchmark is necessary to make a (fair) comparison. I'm very curious to see the vram usage of the game under 8AA (which I've already seen thanks to you ), and curious to see the benchmark result of a dual 580 SLI setup (which I still have no luck to see any)... The only relevant result I've seen before was just a GTX590 getting 34 fps under 1200p at default clock frequency.. If you could run the benchmark at 1080p that would be great - windiff on the log file can tell if we used the same settings or not

bullcrap tall as mt everest lol

the benchmark would be no different.. still double your fps

[sniper_sung]

1280mb is still enough for 1200p 8aa ? hahaha

so why do the game use up to 1500mb out of gtx580 1536mb @ 1080p 8aa ??

I've seen benchmark results of GTX570 managing to be consistent with that of GTX580, without obvious lag spikes there. Worst case? Not even 1.5GB is enough.

bullcrap tall as mt everest lol

the benchmark would be no different.. still double your fps

Wait, what? I knew you dare not show your log file

[NaeKuh]

Super Mario and Toki Dori in a conversation topic of high end gpu's is truely epic... lol...