The OFFICIAL FFXIV Benchmark & Beta Performance Thread

[Celeras]

Unrelated, it feels a lot like I'm capping out the 4870 running at 1776x1000. The gpu is running 99% nearly the entire time (and generally hovering around 30fps until I get near a gathering spot, takes a massive dive). Noticed its also running mostly throughout the low res benchmark, and 99% 24/7 in the high. Upped my cpu .3ghz and saw zero change in scores, and ingame and in the benchmark my cpu never passes 25% (the core focusing on xiv runs about 70% or so). The game is definitely playable, but suffers from AH-syndrome easily with moderate settings on a higher resolution. I think only 512mb is definitely not helping the cause.

Reasonable assumption. For reference sake Purrrfect, my GPU usage is:

-50-70% at 1920x1200 pretty much everywhere in beta.
-99% in benchmark on High, everywhere except the introduction sequence(80-90) and meteor sequence(under 50)
-20-40% in the benchmark on Low

So I finally OC'd my i7 to 3.8 after getting my H50. I ran the Benchmark again, and the only performance I got was an increase of ~200. Any ideas as to why this boost is so marginal?

Because you don't know how to read.

[Kaisha]

Getting about 8-10 fps average on my aging rig, but my CPU seems to be throttling the GPU to the point where it doesn't matter what visual settings I use, it all runs the same. (so I got it all cranked up currently at 1920x1200 >.> )

AMD64 3700+ (2.21ghz) - Single core, she's a relic
GF8800 GT 512mb - Fuck their 9600 requirement, this thing can kick its ass
2gb RAM - Pfft.

Going to try it on my '2 year old MacBook Pro' later to see if that image on the previous page holds any salt as I doubt it'd run slower on mine than it does my PC.

[Zeh]

Excuse me for being illiterate sir celeras, but does ambient occlusion require many resources?

[plue]

Because you don't know how to read.

Well I did read that before, and thanks for linking me back to it. Maybe I misunderstood something and you could clarify it up a bit.

My first test was performed at 2.66 (retail i7 920) speed, at that time I did both low and high tests. The scores were on the lower end of the charts on Page 1.

Currently, I OC'd the processor to ~3.8, which seems like a significant increase. I ran the benchmark again, and on both low and high tests the increase was ~200 points.

After reading what you posted again, I understood that my score should of increased slightly on high res test (GPU intensive), and by a significant margin on the low test.


Even if we forget the high test, since I have two 260's running in SLI, which is not supported right now, and I have not tried to OC them yet. What would cause such a low increase on the low test?

[Myth]

Going to try it on my '2 year old MacBook Pro' later to see if that image on the previous page holds any salt as I doubt it'd run slower on mine than it does my PC.

I'm actually really interested in how the game will run on a MacBook Pro. I'm buying one this week for school and if the game can run on it Bootcamped then I'd be stoked. Be nice to be able to do minor things in-game on a laptop. For when I'd be killing time between work and class.

[Katlan]

http://img8.imageshack.us/img8/4852/1278831407976.jpg

what is this emotion that I am feeling

I'm actually really interested in how the game will run on a MacBook Pro. I'm buying one this week for school and if the game can run on it Bootcamped then I'd be stoked. Be nice to be able to do minor things in-game on a laptop. For when I'd be killing time between work and class.

there you go

[Myth]

there you go

Ah damn yeah I didn't pay much attention to the fact that they had a 2 year old Macbook Pro. Although I wonder how the current generation of Pros compare to that one. I doubt significant improvement, but if you could do things with your retainer then it might be useful.

[Kaisha]

The person with the MacBook Pro has faulty internal fans, and likely never bothered updating their bootcamp with the latest video card drivers, so I'm betting it was underclocking itself to not fry the components.

I'll see for myself in an hour how it runs on mine, still zipping the god damn thing on my PC.

[Celeras]

Even if we forget the high test, since I have two 260's running in SLI, which is not supported right now, and I have not tried to OC them yet. What would cause such a low increase on the low test?

On low, if your score didnt increase dramatically with the OC you have a GPU bottleneck. Run it again and monitor your GPU usage, if its pinned at 99% you could confirm this is the case.

[quannum]

Trying to help a friend figure out the cause for his low scores. His laptop has an i7 Q720, 4 gigs ram, and the GeForce GTS 360M. Other benches in the first post with this processor are scoring about 2000+ on high, while he's scoring just over 1000. The benches from the first post are using the ATI 5870M video card, which from what I can research online should only have about a 10% performance advantage vs his 360M.

Apart from drivers, any suggestions on what might be the bottleneck? Or might this just be another case of the FFXIV benchmark not being particularly accurate?

Appreciate the help!

i basically have the same setup as your friend, Q740, 6gb ram, same gpu. i get 2000 on the low score and can't see the full window on the high setting to see the score

[plue]

On low, if your score didnt increase dramatically with the OC you have a GPU bottleneck. Run it again and monitor your GPU usage, if its pinned at 99% you could confirm this is the case.

Thanks so much for this suggestion. I'll try it again when I get home and get Rivatuner downloaded.

[Celeras]

Thanks so much for this suggestion. I'll try it again when I get home and get Rivatuner downloaded.

http://www.techpowerup.com/gpuz/

Use that to monitor the usage. Doesn't require an install, just download less than a MB (probably) and run. You don't need all the bells and whistles of Rivatuner for this unless you want them for something else

[Sephiroth]

As I understand it, low settings uses more CPU while high settings uses more GPU. So should I expect better performance with my 2.66ghz core 2 duo on high settings with a powerful GPU? It's obviously not a great CPU so it could have issues with a CPU intensive low setting. But it would be weird that such a setup could run better in high settings than low.

[pyroclasmx]

I apparently cant post an image since I don't have 10 posts..... so 5076 on high setting. URL below for screenshot.
pyroclasmx.com High.png

[Celeras]

As I understand it, low settings uses more CPU while high settings uses more GPU. So should I expect better performance with my 2.66ghz core 2 duo on high settings with a powerful GPU? It's obviously not a great CPU so it could have issues with a CPU intensive low setting. But it would be weird that such a setup could run better in high settings than low.

Do people think before they speak at all? It's not going to make your computer magically run better by INCREASING the settings, use your head. There just won't be much dropoff between low/high if your GPU can keep up with the increased settings.

[Zorlin_PT]

Ah damn yeah I didn't pay much attention to the fact that they had a 2 year old Macbook Pro. Although I wonder how the current generation of Pros compare to that one. I doubt significant improvement, but if you could do things with your retainer then it might be useful.

I have the baseline 13" macbook pro mid 2010 with the geforce 320M and i've been playing the beta @1280*800 with most things turn off and screen buffer at half i get 20-30fps. if you dont mind the graphics its perfectly playable. with the 15"/17" models you get the geforce GT330M which is abit better then mine.

[plue]

On low, if your score didnt increase dramatically with the OC you have a GPU bottleneck. Run it again and monitor your GPU usage, if its pinned at 99% you could confirm this is the case.

I just tried your suggestion, alternating between GPU-z and MSI Afterburner to see my GPU loads. Both showed quite similar data, with one card running at ~80% and the other at ~50%. Not sure why it'd be using a load from the second card.

Any suggestions?

[Sephiroth]

Do people think before they speak at all? It's not going to make your computer magically run better by INCREASING the settings, use your head. There just won't be much dropoff between low/high if your GPU can keep up with the increased settings.

I imagine they do, but not everyone is a hardware expert around here. What I had in mind was that in high settings the GPU would take some of the load away from the CPU since the CPU is used less in high settings. In low settings the GPU would be under used but the CPU maxed out. What you said makes sense I guess. I'll find out in a couple of days when my GPU arrives.

[GSDragoon]

Random tid-bit on performance. After installing the beta, there were 84,810 extra fragments in 2.4GB. And since the data is like 11, where the files are smaller individual files as opposted to a few larger files (like steam or wow) I really do think defrags could help with performance.

[Myth]

I have the baseline 13" macbook pro mid 2010 with the geforce 320M and i've been playing the beta @1280*800 with most things turn off and screen buffer at half i get 20-30fps. if you dont mind the graphics its perfectly playable. with the 15"/17" models you get the geforce GT330M which is abit better then mine.

Awesome, I actually just ordered the baseline 13" today. Wasn't looking for anything fancy just lightweight, small, and good battery life. I don't need to kid myself into thinking a laptop is worth $2000 so I can "game" on it. However, if it can run FFXIV on minimum settings I'll take it.