is that good? i must have a play myself. it'll take a few mins to download

I have no idea, I'd imagine it'll be better than yours, but I'm not sure by how much.

3Dmark 11 would probabbly favour your setup due to DX11.

And @ taking a few ins to download.

grabbing vantage atm, about 20 mins left I think.

3 mins to download, half an hour to run it 3 times because 'the benchmark cannot complete 'cause your popups are stealing focus'.

3DMark Score
57852 3DMarks
GT1 - Wings of Fury
590.33 FPS
GT2 - Battle of Proxycon
545.45 FPS
GT3 - Troll's Lair
379.24 FPS
GT4 - Mother Nature
400.54 FPS
CPU Score
1856 CPUMarks
CPU Test 1
170.14 FPS
CPU Test 2
39.09 FPS
Fill Rate (Single-Texturing)
14757 texels/s
Fill Rate (Multi-Texturing)
35780 texels/s
Vertex Shader
154.76 FPS
Pixel Shader 2.0
1035.65 FPS
Ragtroll
205.24 FPS


yeap. bout the same in 03. mines a little bigger, but not so much that i'd inappropriate idiom but i cant think of another. i tested vantage - didnt come out well compared to listings on websites. 14-15k.

Very close, win some, lose some.
Vantage just crashes windows when I try and run, no resolution when searching, cba to try and resolve.

I'll run 03 again in a min with 1 card and see how that runs.

Single GTS 250 1GB on default unregistered.
3DMark Score
39501 3DMarks
GT1 - Wings of Fury
531.6 FPS
GT2 - Battle of Proxycon
341 FPS
GT3 - Troll's Lair
275 FPS
GT4 - Mother Nature
259 FPS
CPU Score
1740 CPUMarks
CPU Test 1
168 FPS
CPU Test 2
35 FPS
Fill Rate (Single-Texturing)
5804 texels/s
Fill Rate (Multi-Texturing)
39037 texels/s
Vertex Shader
172 FPS
Pixel Shader 2.0
516 FPS
Ragtroll
193 FPS


Single GTS / Dual GTS / Single 6850

3DMark Score
39501 | 55864 | 57852 3DMarks
GT1 - Wings of Fury
531 | 479 | 590.33 FPS * Weird that the single card did a lot better than the dual, might retest*
GT2 - Battle of Proxycon
341 | 506 | 545.45 FPS
GT3 - Troll's Lair
275 | 404 | 379.24 FPS
GT4 - Mother Nature
259| 377 | 400.54 FPS
CPU Score
1740 | 1714 | 1856 CPUMarks
CPU Test 1
168 | 163 | 170.14 FPS
CPU Test 2
35 | 35 | 39.09 FPS
Fill Rate (Single-Texturing)
5804 | 11262 | 14757 texels/s
Fill Rate (Multi-Texturing)
39037 | 69286 | 35780 texels/s
Vertex Shader
172 | 187 | 154.76 FPS
Pixel Shader 2.0
516 | 934 | 1035.65 FPS
Ragtroll
193 | 222 | 205.24 FPS

Registered version @ 1680x1050, 8xAA, 16xanisotropic and post processing

Single GTS 250 | Dual GTS 250 - % performance

Score: 9185 | 16110 - 175%
GT1 246 | 381fps - 154%
GT2 46 | 89fps - 193%
GT3 44 | 84fps - 190%
GT4 91 | 156fps - 171%

CPU score 1710 | 1696
CPU 1 166 | 162fps
CPU 2 34 | 34fps

Feature tests
Fill rate 1 5194 | 10168 MT/s - 195%
Fill rate 2 31201 | 61042 MT/s
Vertex 123 | 182fps - 195%
pixel 183 | 341fps - 186%
Ragtroll 33 | 65fps - 196%

so youre looking at about 40% increase for a second card. maybe youre hitting the CPU limit in 03? 75% increase in custom settings. that's good.

your link about battlefield 3 was a nice example of the way things work. a big budget game with a few years development - the devs expect people to have a high-end card along the lines of an nvidia 560 for 2012, to play it at its best, and a good 8800 to play it reasonably well.

they wouldnt spend millions on a game that 99% of people cant play when it's released.

this is part of my problem with the comparisons you see on review sites. they test the cutting edge, using experimental settings that are beyond the scope of both the games and the mainstream hardware. eg, 2560x1600 in ultra settings to give 20fps on a gtx 590. it's a bit meaningless, 'cause most people just won't, and the game devs didnt intend them to. it was possible to run quake at 1600x1200 in the 90's. it didnt make the game look any better – it was the same thing with sharper edges on the blocks that made up the characters, only at 10fps.

id like to see benchmarks with a realistic baseline. the settings used should give 40fps minimum on a mainstream high-end card. ie, the question is, is that £600 card really 3x faster, or are you having to push things so far with unrealistic settings, to show any advantage at all, that the comparisons are void?

itd be fun to spend £400 on a graphics card, to find that as soon as there are games around that wouldnt have played just as nicely on its £100 sister, you cant even run a benchmark to see if it's still worth £60, 'cause the card doesnt support DX 12 : )



Did I win?

Am I right? : )

http://www.youtube.com/watch?v=zYxkezUr8MQ

Morning lappy.

I was agreeing with you for a lot of that, but then you missed the point at the end, some people want the max settings.

I've researched a fair bit on this game now (trying to decide whether to buy it, how well I'll be able to run it and general OOOoooooo, pretty) and read a fair few comments from people saying they want the max settings, else they may aswell just still with a PS3/Xbox (I disagree due to control inputs).

There is currently no hardware commercially available I BELIEVE that could run that game on 3 27" 3D monitors with an average framerate of 120 (min 60) on the ultra settings, except MAYBE the Asus Mars II.

Two of the fastest cards in SLI (GTX 580s) aren't able to run it at Ultra on a 30" monitor resolution...

Yes, I should be able to run it @ an average of 40 (min 18-20 which will be annoying now n then) on High detail (without DX11 stuff), but that's really gonna be good enough for me by a tiny amount. I'd like 25fps min and whatever the average ends up as..

However, I'll know tomorrow (hopefully if Amazon ship the disk today and the post office don't screw me).

BTW, fancy running your card @ my settings on 3dmark, see how they compare?

id like to see screenshots showing ultra vs high. 3 3d monitors and a couple of gtx 580s. id rather have an hd projector with a free ipad 2, or maybe a car.


k. give me a min

I wouldn't use a projector with a res lower than 1680x1050, mostly it'd only be 1920x1080 (because it's a video standard).

Would love to have 3 projectors, but I'd prefer 3 monitors to 1 projector.

3 screens allow for peripheral vision, the side monitors actually angling the view so it is what peripheral would see.
This is most useful for driving / flying games, but also good for FPS'ers.

sounds clever. i'll have a go at that one day if i ever get around to it.

1920x1080, 8AA, 16AF, post proc:

now go and have a shave and a shower, you grubby bastid. or is that me?


**wait, you wanted 1680x1050? hold on, i'll rerun while i take care of the above

Single GTS 250 | Dual GTS 250 | Single 6850

Score: 9185 | 16110 | 21241
GT1 246 | 381 | 426
GT2 46 | 89 | 153
GT3 44 | 84 | 134
GT4 91 | 156 | 159

CPU score 1710 | 1696 | 1843
CPU 1 166 | 162 | 169
CPU 2 34 | 34 | 39

Feature tests
Fill rate 1 5194 | 10168 | 14927 MT/s
Fill rate 2 31201 | 61042 | 38129 MT/s
Vertex 123 | 182 | 161
pixel 183 | 341 | 552
Ragtroll 33 | 65 | 115


Shows how things have moved on.
These cards were 'slightly' slower than the quivilent to yours back when they were released, but now, 3 generations on, it's SOooo much faster.

the 6850 is pretty much what DICE recommend for High settings on BF3, the 8800GT (pretty much the same as these cards) is for low setting.

Guess there is a fair bit more going on between low and high.
Now I'm feeling a little less confident about being able to run it with High preset

DAMN YOOOOUUUuuuuuuuuu

yeah, it's amazing. a couple of years. buy cheap, buy often. that's best advice there is, i think

here, 1680:

i dont think 3dmark 03 is a good test. i doubt it's using half the new stuff, even on a 250

i think 'max' are rarely what game developers intend. im sure it depends on the game. if all you want is the badge to say you can run it in max, and you dont care if it actually looks any different, why not override the application settings completely and run the game in 32x AA? that'd beat it.

ok, if you actually need to turn off features to run in 'high' settings, there may be some difference, but if all ultra is doing is turning off the LOD or increasing the draw distance beyond what's visible... why not just play a video in the background, or buy a cheaper card and spend the difference on having some holograms made that say gtx 690. stick it to the box, add it to the youtube. triple sli gtx 690 battlefield 3 in ultra settings at 40fps. and no one would know. secretly, it's running in normal settings, but it's a quarter as fast as it should be. 'normal'? or ultra? look at the frame rate. for that spec, it must be ultra : )

is the game any good? looks like more of that same old crysis stuff to me.


^ that's a 630 pixelwidth tank right there. im not saying it's huge, but it's not small. im about to test crysis.