The OFFICIAL FFXIV Benchmark & Beta Performance Thread

[alecrast]

Yeah. All three of my HDDs are Seagate 7200 RPMs (1 500GB for backup, 1 320GB for OS, 1 320GB for extra). My CPU is an AMD Athlon II x2 260 Regor clocked at 3.4GHz, and my GPU is an ATI Radion HD 5670 1GB GDDR5. Sometimes my HDDs are slow as heck and I'm not sure if I should just save 100$ and buy a 6GB/s 1TB harddrive and use that for everything.

Unless you're addressing them in parallel, those older 320GB and 500GB drives are definitely going to be slower than a 1TB drive due to areal density - the more tightly compacted the data is over a given area, the faster you can read it. For example, your 500GB drive might have 2x 250GB platters in it while the 1TB drive might have 2x 500GB platters with no change in platter size. What has changed, however, is the density of data on the platters. Not to mention, if you're buying the black edition WD 1TB drive with 6Gb/sec SATA, it's probably the dual-controller model.

With a mechanical hard drive, you're not even going to reach 200MB/sec worth of systained transfer - you might hit 150MB/sec on the outer edge of the platter but that's about it and that's only during sustained sequential reads, something that almost never happens in real-life usage, so 6Gb/sec SATA is really a marketing number. The only home-use drive I have seen to utilize more than a SATA-II (3Gb/sec) connection is the Micron C300 SSD which can hit 350MB/sec sustained sequential reads.

Connecting a SATA-II (3Gb/sec) drive to a SATA-III (6Gb/sec) port can sometimes yield a performance decrease ala the Marvell controllers that are pretty much standard on all new X58/P55 motherboards from Intel. The 6Gb/sec controllers on the AMD 850 Southbridge are hopefully better performing but I haven't really read up on them much.

So just be aware that plugging a SATA-I or SATA-II drive into a SATA-III slot is unlikely to get you any performance increase, if it does, it will be negligible, and the only SATA-III drive on the market that can currently exceed the SATA-II standard is the Micron C300.

[Emirii]

Thanks so much for your help alecrast I suppose I could try defraging and cleaning up the junk of my hard drive too.

[Shaard]

Guys, I'm having a weird issue with the benchmark score going from ~4000 to ~2700 depending on whether the Windows task bar is hidden or not (when not hidden, the bar overlaps a small part of the benchmark screen, lowest score is when the bar is not hidden). Installed OS is Windows 7 (64 bits) and Aero is disabled.

Any idea about what would cause that? I'm not overly concerned, as all is running smoothly enough for my taste, but just out of curiosity, I'd like to understand why this is happening!

If that matters, system specs are the following: i5 750 (stock), nVidia GTS 250 (stock), 4 Gb G.Skill DDR3 1600 (but clocked at 1333 for TurboMode), 1366x768 screen.

Any help would be much appreciated!

[alecrast]

Guys, I'm having a weird issue with the benchmark score going from ~4000 to ~2700 depending on whether the Windows task bar is hidden or not (when not hidden, the bar overlaps a small part of the benchmark screen, lowest score is when the bar is not hidden). Installed OS is Windows 7 (64 bits) and Aero is disabled.

Any idea about what would cause that? I'm not overly concerned, as all is running smoothly enough for my taste, but just out of curiosity, I'd like to understand why this is happening!

If that matters, system specs are the following: i5 750 (stock), nVidia GTS 250 (stock), 4 Gb G.Skill DDR3 1600 (but clocked at 1333 for TurboMode), 1366x768 screen.

Any help would be much appreciated!

Try disabling Aero. The problem is likely that some double rendering is going on. The benchmark is first rendering the benchmark graphics and then, when the Windows taskbar overlaps the benchmark window, Windows Aero is trying to draw the translucent taskbar on top of it.

[Azure247]

I dont know why my score is so low.

AMD Phenom II x6 1090T 3.2ghz
4 GB DDR3
nVidia GTX 480

Score on 1080: 3799

Wont let me post picture because Im new.

[Melion]

Try disabling Aero. The problem is likely that some double rendering is going on. The benchmark is first rendering the benchmark graphics and then, when the Windows taskbar overlaps the benchmark window, Windows Aero is trying to draw the translucent taskbar on top of it.

Aero theme jacks up everything on my media box. Has an Nvidia 7950GT that I can't make get along with it

[HumsterMKII]

I dont know why my score is so low.

AMD Phenom II x6 1090T 3.2ghz
4 GB DDR3
nVidia GTX 480

Score on 1080: 3799

Wont let me post picture because Im new.

Quickly jack up your 1090T to 4Ghz. Should hit 4000+

[Cutriss]

Try disabling Aero. The problem is likely that some double rendering is going on. The benchmark is first rendering the benchmark graphics and then, when the Windows taskbar overlaps the benchmark window, Windows Aero is trying to draw the translucent taskbar on top of it.

Try it, but don't be surprised if it doesn't work. I've read a number of articles where people have actually gotten worse performance disabling Aero, thinking it would actually benefit them. Turns out it's partially tied to some 2D acceleration, so disabling Aero in that case would actually cause the system performance to get worse.

For the record, I haven't seen anyone else with this issue, but I tried to bench my wife's computer a little while ago. The bench "ran" in Low, and I got a final score of 1200ish, but I never actually saw the frames get drawn. The renderer seemed to basically have some viewport issues, as the shading on the logos would be drawn and then never go away, and the utilization graph drew in and never went down.

It didn't run at all in High, it crashed. Was funny to read "We apologize for the inconvenience."

I think it's because the card she has in there now is a 7900 GT with only 256 MB of RAM. I suspect 512 VRAM is a hard minimum for the game. The system itself is a Phenom II X4 920, so based on some other people's results, it should at least be competent, provided there is a suitable GPU card in there (though obviously this didn't help with the Low bench).

[Celeras]

Quickly jack up your 1090T to 4Ghz. Should hit 4000+

At 3800, he's probably being limited by his GPU clock at the moment and won't see much performance increase from his CPU. Just a guess based off the fact that I have the baby brothers of his system (1055T/470) and I wasn't being limited by CPU clock on High until my 470 hit ~750 core.

[HumsterMKII]

At 3800, he's probably being limited by his GPU clock at the moment and won't see much performance increase from his CPU. Just a guess based off the fact that I have the baby brothers of his system (1055T/470) and I wasn't being limited by CPU clock on High until my 470 hit ~750 core.

So you whats actually the ceiling in terms of CPU speed, at High Res?

I got 2700+ at low res and 1700+ at High.

Mine currently is a E7300 which I clocked to 3Ghz. G-card: 4850

When I ran the benchmark, the core usage is mostly 50-65%. But when running at high res, it shoots up to 55-70%+

Sometimes I wonder if the benchmark actually takes note on your G-card, and switch processing to our cpu.

Just a question Celeras, do your CPU usage increase when running at high res, across your 6 cores?

[Celeras]

Ceiling will vary from system to system, clock to clock. My usage is hard to say for sure, one core is usually at 40-75% and the others in the 20-50s bouncing around a lot. It's basically the same as the usage on low though.

[Teryn]

Ran on High with no major adjustments to system other then updating drivers etc. Scored around a 6500 on low but I didn't SS it.


http://img684.imageshack.us/img684/1...mark201007.png

[Shaard]

Try disabling Aero. The problem is likely that some double rendering is going on. The benchmark is first rendering the benchmark graphics and then, when the Windows taskbar overlaps the benchmark window, Windows Aero is trying to draw the translucent taskbar on top of it.

As I said, Aero is already disabled, but your explanation does have its points. I'll try and test some more later, it's just so nagging me right now!

[alecrast]

As I said, Aero is already disabled, but your explanation does have its points. I'll try and test some more later, it's just so nagging me right now!

Sorry, I misread that you had Aero turned off. What driver revisions have you tried so far?

What score do you get if you disable the auto-hide feature of the taskbar so that it is always visible, reboot, and then run the benchmark?

[Rikutheal]

Does anyone know what could be wrong with my benchmark? :S Mine was at 342 on LOW >.<; with a T9600 Dual Core Processor @ 2.8GHz and a 240M Nvidia card. Now considering my computer can run every game under the run, could it be a problem with fraps or other background programs because thats VERY low.

[alecrast]

Does anyone know what could be wrong with my benchmark? :S Mine was at 342 on LOW >.<; with a T9600 Dual Core Processor @ 2.8GHz and a 240M Nvidia card. Now considering my computer can run every game under the run, could it be a problem with fraps or other background programs because thats VERY low.

Update your drivers. Try disabling aero. Try reinstalling Windows cleanly.

Honestly, I'm not going to put TOO much faith in this benchmark since there are a lot of things that can impact a game running in Windowed mode that won't affect it when running full-screen. It would have been very helpful if SE had given the benchmark the capability to run full-screen. So, while your score is low, I wouldn't go into panic mode just yet.

[Shaard]

Sorry, I misread that you had Aero turned off. What driver revisions have you tried so far?

What score do you get if you disable the auto-hide feature of the taskbar so that it is always visible, reboot, and then run the benchmark?

Well, I still don't know why, but now I do know what causes the drop: Windows 7 task bar with small icons. Having the bigger ones doesn't change a thing, only the small ones make the benchmark score drop. Go figure.

S.O. thinks that it is indeed a transparency issue, as you surmised, since the start button is taller than the bar, there must be some weird process to display the bar smaller than it really is (as an object).

Well, in any case thanks for your help and insight! ^^

[Ryn]

Would having those small things make THAT big of a difference for your score?

Doesn't seem like small and big icons would.

[Katlan]

I'm not going to put TOO much faith in this benchmark since there are a lot of things that can impact a game running in Windowed mode that won't affect it when running full-screen.

oh (general question), does windowed mode require more resources then full screen? I had no idea..

[Celeras]

oh (general question), does windowed mode require more resources then full screen? I had no idea..

No