Final Fantasy XIV Benchmark

[Rokudaime Sennin ™]

2728 with i7 @ 3.8 (8 threads) & old 8800GTS 640.
I'll test again after forcing SLI & once more next week when I get a GTX 480.

man, I expected atleast 5000 from yours, that low score with i7? Seriously. wtf do you need to run this game at over 8000, like 4 graphics cards and 16 gb ddr3 ram, man, totally absurd.

[Hand Banana]

[Trigunflame]

Not sure what's causing the low scores.. using newest WHQL 257.21 drivers & forcing SLI is a no-go.

It's obviously not CPU bound though (even at 720p), otherwise I would be blistering everyone else.
I'm averaging roughly ~10-15 overall CPU%, and 4 threads (logical cores) aren't even being utilized.

Second run, again without SLI.
Note: It's showing my CPU down-clocked due to Intel SpeedStep.


And CPU usage, shown on second monitor.

[Hand Banana]

Not sure what's causing the low scores.. using newest WHQL 257.21 drivers & forcing SLI is a no-go.

It's obviously not CPU bound though (even at 720p), otherwise I would be blistering everyone else.

[Rokudaime Sennin ™]

I have a gut feeling that once Trigun installs the new graphics card, and fixes all the drivers, he's gonna rape all the scores, it'll be over 9000

I'm waiting man, hurry it up.

[Smooth Operator]

720p test.


edit: Game is badly optimized. This guy got over 9000 with a 6 core i7 (only 4 cores read?) and an overclocked gtx 480. This was on the 720p test.

[Hand Banana]

Aww dude thats so homosexual. Your processor is just a bit better than mine. I'm rerunning this test.

[Trigunflame]

This guy with a system roughly similar to what I would have, still only got ~4.5k


The guy you posted earlier, likely had a massive CPU OC (not evidenced in this screenshot, he may have said somewhere?), & most obviously had a hefty overclock on his 480 -- specifically on his shaders. (And there in lies the "cock block" of this benchmark IMO).


I'm noticing relatively weaker cards (57xx), matching the performance of 480s on This benchmark & it's likely due to shader processing as there seems to be a Lot of post-processing effects going on during the demo; perhaps more so than I've ever seen in any other game.

Here is an example benchmark of Vantage & Crysis Warhead comparing various cards.
We're looking specifically at the GTX 480 vs the older HD 57xx cards.
(I've omitted older nVidia cards to make it easier to read).

As you can see, a 57xx is handily beaten by roughly 2x by a GTX 480 on both benchmarks.


Yet, on this particular FFXIV benchmark, 57xxs are strangely keeping up with most GTX 480 configurations @ stock clocks. That tells me something weird is going on with either the code paths & optimizations within the program itself; or it's making absurd use of shader processing and that's keeping Older ATI cards up there with the latest Nvidia cards.

So, how might this be possible?

• Well, the 5750 has ~720 unified shaders & the 5770 has ~800.
• The GTX 480 only has 480 unified shaders (although Fermi was supposed to have 512, 1 cluster was disabled).
• And my 8800 GTS g80 (which by the way, SLId & OCd, still runs every game I own @ 1680x1050 res @ 60fps with vsync), only has 96 unified shaders.

That's my theory anyway.

-- Note: I recieved a little over ~100 point increase by disabling Desktop Composition/Themes.

[Hand Banana]

So I ran mine again and it got slightly better. But what was odd is I dropped my performance some on my cpu speed.

[xzero20]

wtf, 8g dd3 ram and quad core and a 1gb vid card, and still only 2500 Wow



I'm afraid to test mine now, I bet I'll barely get 750, lol. 2gb ddr2, intel centrino, intel graphics, LOL.

2728 with i7 @ 3.8 (8 threads) & old 8800GTS 640.
I'll test again after forcing SLI & once more next week when I get a GTX 480.

@ trigunflame daym u gotta be kidding me... even on an i7 3.8ghz 8 threads and ur benching it at 2700... this game looks like its gonna be demanding as hell....

@rokudaime sennin ya.... its pretty ridiculous considering what my baby has. my friend with a slightly older build score a 1978...