The OFFICIAL FFXIV Benchmark & Beta Performance Thread

[shaddix]

Have you tried outside? LL for me dropped GPU usage and slightly upped CPU usage (although it only went to ~25% CPU/~45% GPU, outside was 10% CPU/65~% GPU)

you're right, I get ~60-70% gpu outside and ~60-70% cpu outside. I wish they didn't have that fps cap on it so I would know what the bottleneck is. As it seems inside LL it's the cpu. But the benchmark says it's a gpu bottleneck. I'm not sure which to believe. Is the benchmark useless or is the beta version using an older version of gfx engine?

[Questionable]

Got a Corsair HX850, should be enough to run CF 5870 and prob even 5970 or tri-cf 5870

[Questionable]

you're right, I get ~60-70% gpu outside and ~60-70% cpu outside. I wish they didn't have that fps cap on it so I would know what the bottleneck is. As it seems inside LL it's the cpu. But the benchmark says it's a gpu bottleneck. I'm not sure which to believe. Is the benchmark useless or is the beta version using an older version of gfx engine?

Yeah my cpu fucks up the playability inside LL, it lags like hell, I'm still using the E6550 with the 5830 card.

[Questionable]

What is spam forums?

[Purrrfect]

you're right, I get ~60-70% gpu outside and ~60-70% cpu outside. I wish they didn't have that fps cap on it so I would know what the bottleneck is. As it seems inside LL it's the cpu. But the benchmark says it's a gpu bottleneck. I'm not sure which to believe. Is the benchmark useless or is the beta version using an older version of gfx engine?

You want to get your gpu usage up? Go stand in the middle of some bushes/trees outside LL. I'm wondering if it's all the edges to them works the AA over hardcore. Whatever it is, shrubbery is hard on the gpu.

[Acturus]

You want to get your gpu usage up? Go stand in the middle of some bushes/trees outside LL. I'm wondering if it's all the edges to them works the AA over hardcore. Whatever it is, shrubbery is hard on the gpu.

http://c2.ac-images.myspacecdn.com/i...7da7435901.jpg

[ufo]

If you set the buffer size to double you should be pegged at 99% gpu usage

[Amele]

Do any games use more than 4 threads?

tons of games use more than 4 threads. - this isn't quite the same as saying they can use multiple cores.

crytek engine can if the game is coded properly.

I wouldn't be surprised to find out unreal 3 engine can too.

as for others, maybe metro2033? eve online can thanks to being python64 on the client and multicore friendly, etc.

[Amele]

haha wat 95-100 degrees? 100 is typically the danger threshold for Intel chips and most people don't run them past ~85 for any prolonged period of time. I like to keep mine to ~78- because I'm a vagina.

and to think I try to stay under 73 with mine... (even that's a big jump) tjmax for me is 95, but I'm not really comfortable over 80 for long periods of time because I'm overvolted and don't want to speed degradation...

[Icarian]

Got my new cooler so now I can put my RAM to run dual channel, getting 4800's on high and 5100-5200 on low with a Phenom II X6 1090T @ 4 Ghz and a 5870 @ 940/1290 (shitty RAM might be making me score a lil lower than I should)

How come my low score is so low? there shouldn't be any CPU bottleneck

It's to be expected that the i7's score 200-300 points higher than the x6 1090T at the same speeds, but a 2k+ difference on low? thats WAY too much of a difference D:

[Purrrfect]

Got my new cooler so now I can put my RAM to run dual channel, getting 4800's on high and 5100-5200 on low with a Phenom II X6 1090T @ 4 Ghz and a 5870 @ 940/1290 (shitty RAM might be making me score a lil lower than I should)

How come my low score is so low? there shouldn't be any CPU bottleneck

It's to be expected that the i7's score 200-300 points higher than the x6 1090T at the same speeds, but a 2k+ difference on low? thats WAY too much of a difference D:

This is what I've been wondering about also with my 1055T. Had a 50/50 shot a month ago and went amd... starting to think that might not have been in my best interests.

[Celeras]

How come my low score is so low? there shouldn't be any CPU bottleneck

It's to be expected that the i7's score 200-300 points higher than the x6 1090T at the same speeds, but a 2k+ difference on low? thats WAY too much of a difference D:

Lack of understanding, that's exactly what should happen. Why do you care what your low score is when you'd clearly be playing on high.

[alecrast]

Got my new cooler so now I can put my RAM to run dual channel, getting 4800's on high and 5100-5200 on low with a Phenom II X6 1090T @ 4 Ghz and a 5870 @ 940/1290 (shitty RAM might be making me score a lil lower than I should)

How come my low score is so low? there shouldn't be any CPU bottleneck

It's to be expected that the i7's score 200-300 points higher than the x6 1090T at the same speeds, but a 2k+ difference on low? thats WAY too much of a difference D:

Clock speeds cannot be used to compare x86 processors with different architectures. That comparison broke back in the '90s. The AMD 1055T and 1090T only have a 3MB L2 cache whereas the i7s (non-970/980X) have an 8MB L2. As witnessed by the difference in performance of the i7 980X with its 12MB L2 cache versus identically clocked i7 with 8MB L2 cache, the benchmark loves large L2 caches. It is entirely possible that this does not translate into any difference in playability. In other words, theoretical numbers are purely theoretical until they're proven not to be.

[Picoman]

Here is what i have on the way.

STYLISH BLACK ALUMINIUM TRIGON CASE + 2 FRONT/SIDE USB (Special Offer)
AMD PHENOM II X4 965 (3.40GHz/8MB CACHE/AM3) - BLACK EDITION
ASUS® M4A87TD/USB3: DUAL DDR3,SATA 6.0GB/s, USB 3.0
8GB CORSAIR XMS3 DUAL-DDR3 1600MHz (4 x 2GB KIT)
1GB ATI RADEON™ HD 5850 PCI EXPRESS - DirectX® 11
1TB SERIAL ATA 3-Gb/s HARD DRIVE WITH 16MB CACHE (7,200rpm)
24x DUAL LAYER DVD WRITER ±R/±RW/RAM
600W Quiet 80 PLUS Quad Rail PSU + 120mm Case Fan
SUPER QUIET 22dBA TRIPLE COPPER HEATPIPE CPU COOLER
Genuine Windows 7 Home Premium 64 Bit

Crossing my fingers for a 4000+ score.

[Eladamri]

Is there much difference performance wise simply from wiping an xp 64 bit and going to windows 7 64 bit? or little if none?

What about from 4 gb to 8 gb with this game? For other games I didn't notice much of a difference from 4 to 6 but this could be different..


and Picoman, I have a similar set up with an icooler 5850 turbo, 4 gb of ram and the amd black you mentioned. I just got around 3100 on high, few hundred higher on low. When I used my asus overclocking software to 3.7, I was around 3300 or so. I tried to take it to 4.0, but with my old case right now the thing wouldn't perform. I'm wondering really if it's even worth it to overclock from 3.4 to 3.7 or 4.0 if the difference in game is negligible.

Windows 7 is in the mail along with my case and I was just curious if windows was somehow optimized more for something like a benchmark, or only optimized in terms of the daily things you'd do on your computer. I think I'll get the extra ram anyhow.

[Icarian]

Lack of understanding, that's exactly what should happen. Why do you care what your low score is when you'd clearly be playing on high.

Yeah I'm not too tech savvy, just surprised me to see a 5k low score when people with intels counterpart are getting high 7k's, but alecrast cleared it up for me.

Anyway, like you said, i'ts not that I care about Low scores.

[Questionable]

Is there much difference performance wise simply from wiping an xp 64 bit and going to windows 7 64 bit? or little if none?

What about from 4 gb to 8 gb with this game? For other games I didn't notice much of a difference from 4 to 6 but this could be different..


and Picoman, I have a similar set up with an icooler 5850 turbo, 4 gb of ram and the amd black you mentioned. I just got around 3100 on high, few hundred higher on low. When I used my asus overclocking software to 3.7, I was around 3300 or so. I tried to take it to 4.0, but with my old case right now the thing wouldn't perform. I'm wondering really if it's even worth it to overclock from 3.4 to 3.7 or 4.0 if the difference in game is negligible.

Windows 7 is in the mail along with my case and I was just curious if windows was somehow optimized more for something like a benchmark, or only optimized in terms of the daily things you'd do on your computer. I think I'll get the extra ram anyhow.

300 more on high for windows 7 vs xp. I think due to my XP was overloaded with JUNK. 4Gb should be fine for ram.

Bench means nothing IMO, it feels different in game.

[Gwynplaine]

Is there much difference performance wise simply from wiping an xp 64 bit and going to windows 7 64 bit? or little if none?

I'd like to know more about this too. I use XP right now and have the option of installing win7 64bit. The only purpose I'm interested in changing is gaming. My PC specs, for now, are very moderate and only just meet FFXIVs requirements. Would I see any benefit in switching to win7 64?

Some friends have told me no, I wouldn't. And they also point out that the reqs for win7 are higher than XP. I've seen articles describing benchmarking done that showed winxp scored better. Any thoughts?

[Eladamri]

well the beta currently locks me out because it does not support 64bit xp right now and it's giving everyone problems. I'd upgrade when you're able to.

Curious though about 4 vs 8 gb of ram though

[Picoman]

Is there much difference performance wise simply from wiping an xp 64 bit and going to windows 7 64 bit? or little if none?

What about from 4 gb to 8 gb with this game? For other games I didn't notice much of a difference from 4 to 6 but this could be different..


and Picoman, I have a similar set up with an icooler 5850 turbo, 4 gb of ram and the amd black you mentioned. I just got around 3100 on high, few hundred higher on low. When I used my asus overclocking software to 3.7, I was around 3300 or so. I tried to take it to 4.0, but with my old case right now the thing wouldn't perform. I'm wondering really if it's even worth it to overclock from 3.4 to 3.7 or 4.0 if the difference in game is negligible.

Windows 7 is in the mail along with my case and I was just curious if windows was somehow optimized more for something like a benchmark, or only optimized in terms of the daily things you'd do on your computer. I think I'll get the extra ram anyhow.

Sounds like you have added upgrades to a system over the past few years. Do you think an entirely new build would perform better even though the main parts match your spec?

I tend to run a PC into the ground and then build a new one.