The OFFICIAL FFXIV Benchmark & Beta Performance Thread

[Rasengan]

Just getting a higher "generation" graphics card doesn't make it better. Confusing, I know. An older generation high end card will outdo an entry level newer generation card. For the most part, the numbers to look for in an ATI card to get a better understanding of it's performance are the middle two numbers.

[Toth]

HD 5770 ($~160) or GTX 460 ($~220). What CPU do you have exactly?

AMD Athlon II 630 Quad Core 4 GB RAM

[Toth]

Hmm the 5770 is sold out at BB and since I have my card for there, I was wondering if there are any alternatives there that would be good? There is another version that is $199 for example.

[Jojimbo]

Don't buy the $200 ones, they are overpriced. You can get 768MB version of 460 for that price off amazon.

This is a good priced 5770 (especially for a vapor-X, it's $5 more than the normal version).

[Treize Kordero]

Geez so higher numbers don't mean higher quality, huh?

Okay, what can you recommend for around $100-200? PCI.

For Radeons the higher number states what generation it is. Usually IGP are shit anyway because they won't put all that power on a MB. So a Radeon 4200 might be newer than a 2400 or 2900 but doesn't have the speed or power they have.

If I may, I recommend a Radeon HD 5770. I'm in canada so the prices for them here are under $200. Should be the same if you're in the states. A Radeon HD 5750 is also good if you don't mind the slower GPU and Mem speeds it's also cheaper too but only like $20-40.

If you dont mind spending a few tens (20-30) above $200 then I'd recommend a Radeon HD 5830 which is closer in power to the higher end models. I'd say its the slowest of the 5800 series.

For Nvidia, I'm not a big Nvidia guy but if you prefer them I'm sure a GT240 or GTS250 should do just fine for the same price range. You might want to refer with someone who's more versed in Nvidia cards.

[Toth]

http://www.bestbuy.com/site/XFX+-+AT...201&cp=1&lp=15

How about this guy? I like the amazon one but I need to buy the card from BB since I have a special card for there.

[Treize Kordero]

How about this guy? I like the amazon one but I need to buy the card from BB since I have a special card for there.

That's a decent card.
I'm using a Sapphire Vapor-X Radeon HD 5750 at the moment and it works great. So far I've been able to play Borderlands and Aion at 1920x1080 full graphics using in-game settings without a problem.

I say it's worth it.

[Jojimbo]

Works, is this card a discount thing though? Because you will be paying more for a weaker card otherwise.

[shaddix]

ordered:
GA-890FXA-UD5
GSKILL F3-12800CL7D-4GBECO
AMD Phenom II X6 Six-Core Processor 1055T
CORSAIR Cooling Hydro Series CWCH50-1

current parts:
ocz 850w psu
antec 1200
evga gtx460 1gb
shitload of hdds
dual 21" sony CRTs. lol

gonna sell my tuniq tower(lapped)/q6600(lapped)/e6600/p5n-e sli/gskill ddr2 800/2x 8800gts 640mb(g80)

decent trade no?

[Treize Kordero]

ordered:
GA-890FXA-UD5
GSKILL F3-12800CL7D-4GBECO
AMD Phenom II X6 Six-Core Processor 1055T
CORSAIR Cooling Hydro Series CWCH50-1

current parts:
ocz 850w psu
antec 1200
evga gtx460 1gb
shitload of hdds
dual 21" sony CRTs. lol

gonna sell my tuniq tower(lapped)/q6600(lapped)/e6600/p5n-e sli/gskill ddr2 800/2x 8800gts 640mb(g80)

decent trade no?

I think so. I went on cpubenchmarks to see what you are upgrading from.

From the site:
PassMark - CPU Mark
High End CPUs - Updated 15th of August 2010

CPU Score
AMD Phenom II X6 1055T 5,153
Intel Core2 Quad Q6600 @ 2.40GHz 2,979

I looks good to me benchmark-wise. I also like Gigabyte's AM3 motherboards. I got a GA-890GPA-UD3H, a GA-MA790FXT-UD4P, and a GA-MA785GMT-UD2H. All 3 all beautiful boards work so nicely. :D

I'm hoping you'll find the same satisfaction with yours.

Not sure about the ram I was thinking about getting some GSKILL in the future for the next computer I upgrade but I've haven't had any problems with the Muskhin I've been using.

But I can agree what you ordered is a definite improvement without breaking the bank as well.

[Kylewhataru]

Cpu bottleneck. Runs fluctuate.

I get ~3600 when I lower the clock to 775 (max OC CCC allows), memory is kept at 1100.

So 810/1100 = ~4400
775/1100 = ~3600

If it is a cpu bottleneck the gpu sweet spot is somewhere in between 775 and 810, but how do I get almost 800 more score from only 35 more mhz?

I'm not sure if it's the way CCC handles it or what, I only use afterburner if I want to go beyond 775.

I'm gonna try 775/1100 w/ afterburner and no CCC

Edit: Yeah, it's CCC, got 4401 @ 775/1100 with afterburner. What would cause this? Exact same speeds just using different apps. The only thing I can think of is CCC throttles the cards down to conserve power, whereas afterburner keeps it running at your set speed 100% of the time, but that shouldn't affect performance besides a second of stutter when you start a 3D application while it adjusts from idle to load speeds.

[Toth]

Okay picked up the card and installed it but now I have a new problem. The monitor cable I have is not compatible with this new video card but works fine with the old one. I have to disconnect the new card for the monitor to come on (and use the old card). Is there a way to configure it so that it will read the monitor input through the old card or do I need to buy a specific cable?

My monitor has a DVI-D connector port. Is the necessary cable expensive?

[Kylewhataru]

It should've come with VGA/DVI adapters, if it didn't it's only like $5.

Looks like this http://www.tvcables.co.uk/images/items/vga-to-dvi.jpg

[Akiyama]

I get ~3600 when I lower the clock to 775 (max OC CCC allows), memory is kept at 1100.

So 810/1100 = ~4400
775/1100 = ~3600

If it is a cpu bottleneck the gpu sweet spot is somewhere in between 775 and 810, but how do I get almost 800 more score from only 35 more mhz?

I'm not sure if it's the way CCC handles it or what, I only use afterburner if I want to go beyond 775.

I'm gonna try 775/1100 w/ afterburner and no CCC

Edit: Yeah, it's CCC, got 4401 @ 775/1100 with afterburner. What would cause this? Exact same speeds just using different apps. The only thing I can think of is CCC throttles the cards down to conserve power, whereas afterburner keeps it running at your set speed 100% of the time, but that shouldn't affect performance besides a second of stutter when you start a 3D application while it adjusts from idle to load speeds.

Wait. Afterburner causes you to run your clock speeds 100% of the time and doesn't underclock when it doesn't need as much of an 'umph'? I know when I'm playing games that don't need as much 'power' (in terms of core clock and what not), it underclocks to around 324MHz. Like with PSO: Blue Burst -- with a 30 FPS cap, it doesn't need that much at all (like 12-20% usage on a GTX 460 lol), so it underclocks.

That is if I'm reading what you said correctly. D: I personally hated CCC.

[Cooleko]

Played FFXIV Beta today for over 2 hours, everything maxed out in the config and in game all options on high, shadows etc.

Im just perplexed. People saying RAM seems to be at fault bad over clocks etc yet games are playing flawlesly. I guess thats what really matters? but i dont want it to just randomly explode one day.

I did the BIOS reset and the CPU seems to be still overclocking, itself maybe? Its supposed to be 2.67 GHz as clock but its showing up as 2785 MHz (21 x 133) What would cause this? Does anyone have any idea why CPU-Z, Everest and other monitors are picking up my CPU at running this overclocked speed yet when i boot up the machine the initial loading window picks it up as 2.67 GHz?

According to the RAM (http://www.ocztechnology.com/product...triple_channel) it is supposed to be CL 8-8-8-24 (CAS-TRCD-TRP-TRAS) yet the monitors show it up as 7-7-7-18 (CAS-TRCD-TRP-TRAS) DRAM 530.6~ MHz NB Frequency 2122.6~ MHz

This doesnt make sense as i did the Bios reset so why are the timings off still?

You have your bios settings set to AUTO, change them to the values instead of auto.

[Kylewhataru]

Wait. Afterburner causes you to run your clock speeds 100% of the time and doesn't underclock when it doesn't need as much of an 'umph'? I know when I'm playing games that don't need as much 'power' (in terms of core clock and what not), it underclocks to around 324MHz. Like with PSO: Blue Burst -- with a 30 FPS cap, it doesn't need that much at all (like 12-20% usage on a GTX 460 lol), so it underclocks.

That is if I'm reading what you said correctly. D: I personally hated CCC.

If you use it to OC past CCC's limits, yes, CCC limits you at around 10% OC, so that's the point of using afterburner. But afterburner wont override that limit unless you go to the cfg file and change one of the options from 0 to 1, that has the effect of disabling powerplay (the energy saving/throttle feature) as well.

This is with ATI cards, I assume it doesn't affect nvidia cards since the option I'm talking about in the cfg says ATI.

"ATIADLHAL
EnableUnofficialOverclocking = 1"

That's what it looks like.

At this point I'm pretty sure the benchmark just doesn't handle powerplay well. In every other benchmark I run the difference between powerplay on/off is only like 2-3%, in ff14 benchmark it's over 20% difference in score (3600vs4400, same exact clocks). When running the benchmark w/ powerplay on I can see when it's throttling up, it's very obvious in the the 2nd part in the boat, it starts out very simple, just some rain and only a few characters rendered, then that jellyfish swarm suddenly rushes in and and I can see the stutter as the card throttles back up.

I also remember people saying that in the beta that they had that stutter with ATI cards when they loaded into new areas, stutters for a few sec then goes smooth, that is the card throttling up.

[Arximiro]

Here is my score with my new card.

Cpu@ 3.0
Radeon HD 5770 1GB GDDR5 @ default Speeds
8GB RAM @ default speeds

http://i36.photobucket.com/albums/e4...-cardStock.jpg

For some reason my load time is higher, and this is the second time in a row I ran it, so it didn't cache it I guess. I'm happy with this score, runs the game fine which is all I wanted. ^^

P.S. Not going to run one on high, everytime I do it stretches onto my second monitor and it's annoying, and my monitors are maxed at 1680X1050.

[Akiyama]

If you use it to OC past CCC's limits, yes, CCC limits you at around 10% OC, so that's the point of using afterburner. But afterburner wont override that limit unless you go to the cfg file and change one of the options from 0 to 1, that has the effect of disabling powerplay (the energy saving/throttle feature) as well.

This is with ATI cards, I assume it doesn't affect nvidia cards since the option I'm talking about in the cfg says ATI.

"ATIADLHAL
EnableUnofficialOverclocking = 1"

That's what it looks like.

At this point I'm pretty sure the benchmark just doesn't handle powerplay well. In every other benchmark I run the difference between powerplay on/off is only like 2-3%, in ff14 benchmark it's over 20% difference in score (3600vs4400, same exact clocks). When running the benchmark w/ powerplay on I can see when it's throttling up, it's very obvious in the the 2nd part in the boat, it starts out very simple, just some rain and only a few characters rendered, then that jellyfish swarm suddenly rushes in and and I can see the stutter as the card throttles back up.

I also remember people saying that in the beta that they had that stutter with ATI cards when they loaded into new areas, stutters for a few sec then goes smooth, that is the card throttling up.

Oh, okay then. Yeah, this is on my NVidia card, so I wouldn't know. Best ATI card I had was a 4850 and there's no way I was going to OC that thing D:

[Questionable]

Anyone have issues with being in town, there's massive lag spikes, not constant lag but certain areas and view directions makes my system freeze up every few seconds.

This is why a i7 930 and 5870 - so W T F?

[shaddix]

Anyone have issues with being in town, there's massive lag spikes, not constant lag but certain areas and view directions makes my system freeze up every few seconds.

This is why a i7 930 and 5870 - so W T F?

hdd maybe?