Whats your FFXIV Benchmark score?

[mindless728]

probably not, the 8800 Ultra isn''t that much faster, i would go something like ATI 5770, decently faster and has low power consumption (8800 Ultra is a few generations old now)

 

[alextheawesome]

I just did a male Hyur high one. 4258.

 

[fryhtaning]

The stats I looked up showed the Ultra with 50% more in just about everything - Bandwidth, flops, pixel fill, and texture fill. Also about 25% faster in all the clocks.

http://www.gpureview.com/show_cards.php?card1=475&card2=515

I guess if it flops, I'll just slough it off onto my wife and have a look at the HD 5770. My concern though is that the stats seem to be not much higher than the 8800 Ultra, and that's at PCI-E 2.1 whereas I only have 1.x. I have 700 watts and plenty of cooling so I'm not too worried about consumption.

I guess my question would be - what stat does the 5770 have that would blow away the 8800 Ultra on a 1.x mobo? It's been a while since I used to be savvy in GPUs...

 

[mindless728]

The stats I looked up showed the Ultra with 50% more in just about everything - Bandwidth, flops, pixel fill, and texture fill. Also about 25% faster in all the clocks.

http://www.gpureview.com/show_cards.php?card1=475&card2=515

I guess if it flops, I'll just slough it off onto my wife and have a look at the HD 5770. My concern though is that the stats seem to be not much higher than the 8800 Ultra, and that's at PCI-E 2.1 whereas I only have 1.x. I have 700 watts and plenty of cooling so I'm not too worried about consumption.

I guess my question would be - what stat does the 5770 have that would blow away the 8800 Ultra on a 1.x mobo? It's been a while since I used to be savvy in GPUs...

this chart has the 8800 Ultra and the GTX260 (which is as fast as the 5770)
http://www.tomshardware.com/charts/gaming-graphics-charts-2009-q1/Left4Dead,1182.html

GTX260: 93.40 FPS
8800 Ultra: 75.00 FPS

and the 5770 is only $150 card right now

 

[fryhtaning]

this chart has the 8800 Ultra and the GTX260 (which is as fast as the 5770)
http://www.tomshardware.com/charts/gaming-graphics-charts-2009-q1/Left4Dead,1182.html

GTX260: 93.40 FPS
8800 Ultra: 75.00 FPS

and the 5770 is only $150 card right now

But how does that change on a 1.x PCI-E motherboard? I thought 2.0 loses a significant amount of bandwidth on 1.x.

 

[mindless728]

But how does that change on a 1.x PCI-E motherboard? I thought 2.0 loses a significant amount of bandwidth on 1.x.

PCI-E 1.X x16 slot is the same bandwidth a PCI-E 2.X x8 slot, and even the 5870 (read about twice as powerful as the 5770) only loses something like 4% so the 5770 probably won't lose anything or be a negligible loss

 

[Kraze7]

I think midless728 is correct frythaning. My old 8800 GTX Superclocked(RIP) got high around 2300. Also have you tried to OC your CPU even a little? I think a lot of people with Q6600 can get over 3.0Ghz easily.

I have reached the limit of my GTX 480 with stock fan cooling. Until I get Water Cooling anything above these settings my system crashed.

Core Clock: @830
Processor Clock: 1665
Memory Clock: 2065
Fan Speed: 100%
Temperature capped: 84C~ Full Load.

Result?
On Low:

 

[fryhtaning]

PCI-E 1.X x16 slot is the same bandwidth a PCI-E 2.X x8 slot, and even the 5870 (read about twice as powerful as the 5770) only loses something like 4% so the 5770 probably won't lose anything or be a negligible loss

Very enlightening! Thanks!

My wife has an nForce board and I have an intel 975x... maybe if the score is not impressive, I could give her the 8800 Ultra and crossfire 5770/5870 on mine? Then if she's still lacking we can just SLI hers since it's an nForce board. Sound like a plan?

 

[Kraze7]

fryhtaning,

At the moment SLI or CF won't help improve your score because the benchmark only using a single GPU. My hope is that when the game release in Sept? it will support SLI/CF. But you can test and see for yourself. Take it slow and don't rush.

 

[mindless728]

i guess that could work, it also depends on what you are happy with as well, the 8800 Ultra may be enough (i have 2 x 4870's which is the same as 2 x 5770's and sometimes i think about upgrading to 2 x 5870's)

 

[fryhtaning]

fryhtaning,

At the moment SLI or CF won't help improve your score because the benchmark only using a single GPU. My hope is that when the game release in Sept? it will support SLI/CF. But you can test and see for yourself. Take it slow and don't rush.

I can see how it came across that way, but I really didn't mean that the crossfire would help my SCORE, per se. But considering all the previews into FFXIV having multi monitor support, it's probably a surefire bet that it will support it. Of course it only makes sense to buy the first one right away.. for now

 

[fryhtaning]

From what I've been reading, I can't decide between a single 5870 or xfire 5770s. If I go with the former, I should wait until right before release date in order to let the price drop a little more, otherwise there's no worry about jumping on the first 5770 up front.

 

[GenTarkin]

There is a method to run the benchmark in fullscreen but it does not work for all.
I tried it and was fairly successful. My monitor does not support the high resolution setting so only have fullscreen result for SLI of 250GTS on low, with a i7 920 @ 3ghz .99v HT off

Non SLI ( 1 GTS 250 ) low = 3650
SLI ( 2 GTS 250 ) low = 5560

The method uses a hacked DX9 dll and runs the benchmark .exe name of 3dmark06....so it tells nvidia drivers to run the benchmark using 3dmark06 SLi profile.

 

[Kraze7]

fryhtaning, it's ok to wait for now at least since the game won't be out in two months or so. But once you do get the new card(s) make sure you have time to test run them. The last thing you want is to have a new non working card and have to do the RMA process. The collector edition will get to play few days before sd edition.

PS. When I bought the GTX 480 it was on a weekend sales -$20 and on top of that -$30 rebate. From $504.99 So $50 off was a nice deal. Two free games lol.





GenTarkin, very interesting that SLI GTS 250 can scores 5560 on low. Is there any way for you to test it on high? I understand that your monitor doesn't support high res but maybe borrow a monitor from a friend or something =)

If anyone with SLI/CF can also test the with this method I think it would be nice to see the results. And also GenTarkin, maybe you can explain a bit more about this DX9.dll hacks for benchmark?

 

[rmerwede]

Low: 5292 / 17137ms

High: 4365 / 17238ms

26" 1920 x 1200

specs in profile...

HD 5850 reached 67c max @ 99% load

Looks pretty good

 

[Rustyy117]

No one beaten my high score yet?

 

[misterx1969]

OK damn near the same video card, identical ssd boot drives (sata) in both Patriot 32 Gb. Both using Old 32 Bit Windows XP on Multicore machines. Dual AMD beats Quad Intel on CPU intensive? Both these tests on low. Dont know how much that is really bottlenecking, only on 1 PCI 16x 8600 GTS in each machine, think the non silent has more memory?

1128 LOW

on dual core athlon xp (low power model) with SSD boot drive (Benchmark NOT on SSD)
2x 2.0GHZ 1 GB ram On Windows XP 32 Bit with Nvidia Geforce 8600 GTS with Fan

http://www.facebook.com/editalbum.php?aid=2023150&add=1&flash=1#!/photo.php?pid=30648697&id=1595649890&fbid=1299317175104

Sorry TH cant get photo links to work but there it is

NOW the interesting thing..

811 LOW

on quad core intel with SSD boot drive (Benchmark on SSD)
4x 2.4GHZ 2 GB ram On Windows XP 32 Bit with Nvidia Geforce 8600 GTS silent

http://www.facebook.com/editalbum.php?aid=2023150&add=1&flash=1#!/photo.php?pid=30648698&id=1595649890&fbid=1299317215105

 

[GenTarkin]

http://www.psu.com/forums/showthread.php?t=235360

Refer to the above link to attempt full screen hack method. Unfortunately I dont have a friend with a higher res monitor. Sorry =(

Also, what the above link does not mention is, when you first run benchmark the config.ini is created, in resolution entry in that file, 0 is for 1080p 1 is for 720p

 

[notty22]

a gtx 460 preview......

 

[Niklas_13]

low 5472/17605

high 3963/28731

I5 750 stock ATM HD 5850 stock 24" asus 1920x1080 winblows 7 64 bit

 

[Niklas_13]

from when did i5 750 3.4ghz became stock.How u manged 5472 with 5850.my 5850 725/1000 and q9550 2.83ghz managed 4502

My CPU and GPU are stock atm i have not changed my sig so it shows what my CPU was when i OCed it. really not to hard to figure out is it ?

Anyways maybe my i5 750 is better then the 9550 in this benchmark

Ran again today

low 5355 / 46128 low 5298 / 17665

high 3948 / 17493

Maybe differnt drivers also ? CCC 10.6

 

[Niklas_13]

can u post a gpuz screenshot of ur card.and also the benchmark screenshot.

sure let me run them again and post

 

[Niklas_13]

http://img203.imageshack.us/i/benchmarkhigh.jpg/

http://img686.imageshack.us/i/benchmarklow.jpg/

http://img684.imageshack.us/i/gpuc.jpg/

there you go

 

[Niklas_13]

It seems the game likes i5 system.becoz if u see the first page of this thread u will ppl with 5870.5970.gtx 285 scored lower than urs.and in another thread i saw a person getting 42XX score with 5850 and i7 920 stock.

shurgs just posting what score it pops out

 

[Kraze7]

No one beaten my high score yet?

Lol probably not with the stock cooling. But I did get some more improvement on both low/high. There is someone getting 8000+ on low with i7 980X / 5850. Low setting testing more of the CPU usage.




High around 4800~4900.
Low. upped from 7000 now to:

Really came a long way lol and I'm happy with my results.

Trail #1.
High: 4,244 @1920x1080 | Load Time:10781
Low: 6,250 @1280x720 | Load Time:10808

Trial #2.
High: 4,707 @1920x1080 | Load Time:10446
Low: 6,822 @1280x720 | Load Time:10979

Trial #3.
High: 4,696 @1920x1080 | Load Time:10425
Low: 6,976 @1280x720 | Load Time:10580

Trial #4.
High: 4,700 @1920x1080 | Load Time:10485
Low: 7,054 @1280x720 | Load Time:10480

Trial #5.
High: 4,800~ @1920x1080 | Load Time:10488
Low: 7,201 @1280x720 | Load Time:10448