• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

FFXIV Benchmark

720:
4013 score
11329 load time

1080:
2794 score
11168 load time

not bad I suppose for an old system (sig rig). Didn't utilize the cores like I would hope as I barely crossed 50% on all four at any given time. Didn't really monitor the vid card to see it's utilization.
 
That's not even the half of it. The 5850 trounces the GTX 470. :(


Well it's not like I'm going to play this I think. (Still play WoW.) But still, it's the principle.


my guess is because its a dx9 game only and the 400 series is setup for dx10/11 performance on not dx9 performance.. thats just my guess.. but then again its a friggin benchmark and not the actual game..
 
Thought I'd add some thought on this, I got into the Alpha build my Rig (below) scored 4907 on 1080P haven't scored 720p.

This game ruined my Rig I'm chalking this up to it being the Alpha...but we're talking barely 20FPS at 720p (couldn't change the res in the Alpha Build) this boat scene was instance based and is the opening of the game to those wanting to know. Also the game is capped at 30FPS...which I rarely...if ever saw, I seriously don't know how this game will run on Eyefinity or 3D Surround setups based on what I played.

Again...it was an Alpha, The Beta goes live July 17th I'll let it be known if the performance is improved by then.

EDIT--
Wanted to add that there is a way to make the game run on Multi GPU (at least the benchmark) there's a youtube video of someone doing it by renaming the EXE to 3DMark06 so for you guys with Multi GPU setups go ahead and try that to see if your score improves, also turn off Aero, that alone netted me 350 points.
 
Setup in sig 1080p:
10.4 drivers: 1440
10.6 drivers: 1501

Wow, this benchmark makes me sad when a 512m 4870 beats me. No other issues in games or performance so shouldn't be any driver problems.
 
New CPU new score.

Adding 2 extra cores added 1500 to score and cut the load time in half.
 
Wanted to add that there is a way to make the game run on Multi GPU (at least the benchmark) there's a youtube video of someone doing it by renaming the EXE to 3DMark06 so for you guys with Multi GPU setups go ahead and try that to see if your score improves, also turn off Aero, that alone netted me 350 points.

Anyone try/confirm this yet?
 
I thought I'd shed some light on the current scores.

this benchmark seems to be either based off of the alpha or early beta client of the game, which currently still needs to be optimized quite a bit. so far it seems that all optimization testing SE has done is based off of the intel cpu because amd's cpus' performance is quite noticeably lower in most equally matched cpus (something that hasn't yet been brought up on the alpha/beta feedback forums, since the benchmark has been the only indicator).

currently, I'm running a q6700@3.2ghz with an afterburner oced 5770 on xp 32-bit (I know, no dx11 support, but I plan to transfer the card, lol), and reaching scores of ~4400 (low) and ~2900 (high).

for those of you on vista and 7, disabling aero can increase performance up to a few hundred points (not quite as noticeable as it is in the client, many reported the game running much more smoothly after disabling it).
 
That makes sense...
in a recent interview at e3, game producer Hiromichi Tanaka even mentioned that there was still optimization to do and they they would be getting with Intel to do so. No mention of AMD at all.

As a side note, upgrading from an HD4870x2 to an HD5870 supercharged my performance in the FFXIV bench (among many other things). Low score went from 3000ish to 7000ish, high score from 2500ish to 5000ish. Very pleased with the boost.
 
FFXI would only run at a few odd resolutions internally and would then be upscaled to your selected resolution - just like many console games. I hope this new game can run natively at any selected / supported resolution, and not just 720p / 1080p.
 
turning off areo does not make a difference in my scores so that one's busted

renaming exe did nothing as well so it would appear that there are no cheats lol

Same here. Only thing that made significant differences for me was turning OFF SLI improved my scores.
 
FFXI would only run at a few odd resolutions internally and would then be upscaled to your selected resolution - just like many console games. I hope this new game can run natively at any selected / supported resolution, and not just 720p / 1080p.

The only internal resolution portion of FFXI was the texture resolutions you could pick, instead of offering us things like 'low/med/high/ultra' they had 256x256, 512x512, and 1024x1024, and if you were adventurous or weren't happy with that, registry edits would render the textures at even higher resolutions.

It might have run at weird resolutions on the consoles, but on the PC it was the texture resolutions and the UI elements that were scaled by the user independently of the screen resolution.
 
1080p

2 x GTX 480 SLI @ 810/2116, core i7 @ 3.6ghz, 12GB DDR3 1600 @ 8-8-8-24, 3x X25-M 80GB RAID0

257.21 drivers

Score: 4553

Just upgraded to an i7 980x the other day, clocked it at 4ghz, and boosted the clocks on my 480s. (Rig now in sig).

Now I get 5011 points
Loading time was just barely above 10000ms
 
The bench is not SLI/Crossfire aware, so you're getting the points you'd get with one card, not two.
 
The bench is not SLI/Crossfire aware, so you're getting the points you'd get with one card, not two.

Yeah, I'm aware. I noted that in my original post on the first page but I snipped it out of my quote because it was redundant at this point :p. Still, it's quite strange to release a benchmark that doesn't support it.

Also a little disappointing on the Nvidia side of things considering the whole TWIMTBP business it has going. Not that I plan on exchanging all of my free time for this video game!

Edit: I disabled SLI and ran the benchmark again and barely even broke 4000. That's 1000 points less than having SLI enabled. Anyone else get that?
 
The only internal resolution portion of FFXI was the texture resolutions you could pick, instead of offering us things like 'low/med/high/ultra' they had 256x256, 512x512, and 1024x1024, and if you were adventurous or weren't happy with that, registry edits would render the textures at even higher resolutions.

It might have run at weird resolutions on the consoles, but on the PC it was the texture resolutions and the UI elements that were scaled by the user independently of the screen resolution.

Thanks for clearing that up - all this time I thought it was actually rendering the image offscreen at those odd resolutions!
 
i5 @4ghz, shows 3.8 on bench tho. XFX 5870 v2. 1080p. Above is a ridiculous score for 5850s.

ffxiv.jpg
 
combing through the results the ati cards don't seem to load faster than 14,000ms

gf8/gf9/gf200 get 9,900-13,000ms load times

one guy pulled off 10,000ms with a gf400 but others were all above 14,000ms.

I've tried a i7-920 @4.0ghz fresh install of windows 7 64bit on a ocz vertex 2 and got 20,000ms load times with task manager peaking the exe at 89%. hd5450 (crappy card)

3.2ghz e8400/8600 w/ an 8800gtx 8gb of ram got 12,000ms
 
Meh, one of my 5870s got 4859 @ 1080P. I had Crossfire off. Haven't tried it on my OCed 470s yet.
 
Last edited:
i5@4ghz w/ ati 5870 875 core/1250 mem 10.6 catalyst 1080p

ff_bench.jpg
 
having played the beta last night it, doesn't matter what you got. The game runs like shit. Hopefully they clean it up. TERA performs much better surprisingly.
 
Yea it performs nothing like the beta. Something is wrong there. I know it wont be completely accurate, but it should be somewhat close... Its not.
 
As another beta tester, yes I completely agree it's ridiculously awful. You should try playing at 5760x1200 it's just a slideshow with everything turned off.
 
Yea eyefinity is laughable. It was less than a slideshow for me. Even ran kind of shitty in 1920x1080. Too bad benchmark isn't somewhat accurate. I could be playing in eyefinity at playable settings.
 
720:
4013 score
11329 load time

1080:
2794 score
11168 load time

not bad I suppose for an old system (sig rig). Didn't utilize the cores like I would hope as I barely crossed 50% on all four at any given time. Didn't really monitor the vid card to see it's utilization.

I'm running a Q6600 at 3.2, and a 2GB 5870 and I'm getting 4031 at 1080p, I don't think it's the game not utilizing all 4 cores that's causing your poor performance, it's the videocard
 
Core i7 920 at stock 2.66ghz
SLI GTX 480's at stock 700mhz / 3.7ghz RAM
80GB X-25M G2

ffxivbench.png
 
Has anyone been able to get that fix working with SLI? I downloaded the DLL, put it in the folder. FF14 benchmark runs at fullscreen now but Precision reports 99% load GPU1, 0-4% GPU2 for the duration. Also, my score is still ~4900. Load time didn't change, still ~10300.

*I renamed the benchmark to 3DMark06.exe, Crysis.exe, Bioshock.exe, etc... nothing works.
 
Last edited:
Has anyone been able to get that fix working with SLI? I downloaded the DLL, put it in the folder. FF14 benchmark runs at fullscreen now but Precision reports 99% load GPU1, 0-4% GPU2 for the duration. Also, my score is still ~4900. Load time didn't change, still ~10300.

*I renamed the benchmark to 3DMark06.exe, Crysis.exe, Bioshock.exe, etc... nothing works.

I don't know about all that, but I can confirm it works in Surround, which I read is a component of SLI. I had to turn off Ambient Occlusion, Depth of Field and turn down shadows, so I'm pretty sure it was only rendering on one card.
 
Ok, I got it to work. But holy shit nvidia, get your shit together. Benchmark says my i7 is at 2.9 but evga-eleet reports 4.0. This is on "high". Literally any CF configuration will massacre SLI. I hope the release is better.

Score: 6333
Load: 10538ms

http://imgur.com/a6HGz.png
 
Wow, thought I had a half-way decent system.

At 1080p, I had a score of 3,679 with a q9550 at 2.8 (stock) and a 5850 at stock speeds.
 
Back
Top