• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

*** OcUK Crysis DX9 TopList ***

Very true across the board Greebo .

I have 2 identical systems and they both act totaly different .

1 system will clock up to 4.5 in 100mhz steps and be stable all the way and the other dosnt like anything but 3.0 or 4.0 ,, no other settings are stable .

One of my 8800gtx's will clock to 645 and the other hangs at 620 max even tho they were bought at the same time and have concurrent serial numbers .

I guess that when you put billions of transisters in a computer case there is bound to be variances .
 
SimonMaltby said:
Hi all... been awhile since i looked in, mainly cos i got rid of all my kit and changed to an M1730 lappy so my overclocking days are on hold for now.... Cant believe i'm still on top spot :eek:

RavenXXX2.... Couple of comments i would make.
Firstly when the 9800GX2 first came out there were poor drivers but some very good beta one's which i'm sure helped get these scores. Also there have been patches for Crysis which may well have fixed some bugs and the pre patch may acually have better performance (i'm guessing here).

What i can say is this. I took a totally different approach to most overclockers on the setup where i got great benchmarks...

Firstly i kept the FSB at 333Mhz when booting and just made minor adjustments in windows - this gave me much better control over every aspect of the setup from ram to pci-e speed. I'm afraid i can't remember the exact settings, but i spent many hours perfecting the right balance and i do remember that it made a great deal of difference.

The bigest advantage of not usin high FSB is that you can keep the NB and SB cooler.

Secondly, i took a great deal of time and attention to the OS. Using server 2008 64 Bit with only minimal installed components helped a huge amount. Stopping as many processes and services (including the clock) also produce gains.

Anyway. I asure you the benchmarks are genuine. I was able to re-create within a few fps although of course the setup was far from stable for real use at those speeds.

Come on someone... knock me off the top... I cant say i deserve to be there when i dont have the kit anymore.
Click to expand...

So you overclocked everything from in Windows??
 
Big.Wayne said:
Yip he's deffo got issues!
Click to expand...

oi.... :mad: :(

Would love to see some 1680x1050 or 1920x1200 results from actual users but this database hasn't been configured to allow anything higher under WindowsXP :confused:
Click to expand...

that form is for my dx10 only very high settings thread....

http://forums.overclockers.co.uk/showthread.php?t=17858884

beh* said:
Hi gurusan, can you please update the list to reflect my benchy?
Click to expand...

i'll answer this.... you need to submit your score via the form in the first post. gurusan doesn't trawl the thread looking for new scores. he just posts a database dump from the scores submitted...... :)
 
willhub said:
So you overclocked everything from in Windows??
Click to expand...

Not everything no.

CPU multi and voltages are set in Bios... so are ram ratio and times.
I changed FSB, Pci-e fsb and GPU clocks in windows.

My method was only possible because i splashed out on the multi unlocked cpu of course
 
marc2003 said:
oi.... :mad: :(
Click to expand...
Hey don't worry about it, it's only a few fps and as u said most likely something to do with Vista?

I think your database skills are very good and your work is very much appreciated, I wonder does it use a lot of your personal resources?

I'm not so concerned about the competitive element to the top list but I thinks its a very important and useful reference for a lot of folk. I wondered would you consider extending the function to allow for higher resolutions (1680x1050 or 1920x1200) under WinXP as we have now had some next gen hardware that can achieve playable framerates at higher res?

Thanks for all your hardwork and hopefully u may realise I am a nice guy and stop giving me gip! :cool:

Hopefully that should do the trick, thought I put that nicely?
 
SimonMaltby said:
Not everything no.

CPU multi and voltages are set in Bios... so are ram ratio and times.
I changed FSB, Pci-e fsb and GPU clocks in windows.

My method was only possible because i splashed out on the multi unlocked cpu of course
Click to expand...

Does changing the PCI-E freq really help? And what software you use?
 
Big.Wayne said:
I'm not so concerned about the competitive element to the top list but I thinks its a very important and useful reference for a lot of folk. I wondered would you consider extending the function to allow for higher resolutions (1680x1050 or 1920x1200) under WinXP as we have now had some next gen hardware that can achieve playable framerates at higher res?

Thanks for all your hardwork and hopefully u may realise I am a nice guy and stop giving me gip! :cool:

Hopefully that should do the trick, thought I put that nicely?
Click to expand...

check out my submit page. it's been updated.... :p
 
Concerning all the differing framerates on *similar* hardware, what version/patch of Crapsis is everyone using to run their benchmarks with?

I've currently just been using the retail v1.2 patch, my max score is still ~52 ish under WinXP on this rig, and about 2-3 fps less on Vista/2K8. I also didn't see any noticeable gain when going from 800Mhz to 1066Mhz DDR2 RAM.
 
Skyboat said:
Concerning all the differing framerates on *similar* hardware, what version/patch of Crapsis is everyone using to run their benchmarks with?

I've currently just been using the retail v1.2 patch, my max score is still ~52 ish under WinXP on this rig, and about 2-3 fps less on Vista/2K8. I also didn't see any noticeable gain when going from 800Mhz to 1066Mhz DDR2 RAM.
Click to expand...

Running the 1.2 patch too.
 
Personally, i think it's time to move on now, from 1280x1024 resolution. Most people will have 1680x1050 i believe (and quite a few 24" 1920x1200 resolutions)

As this is a "DX9" test, i suggest it be kept that way. However, how about doing a benchmark of, (say) 1680x1050 - with possible "AA/AF"?. There is already a DX10 benchmark of course. Reason I say "1680x1050" res, is that these resolutions can be run with 20", 22", adn 24" monitors - and above, of course. But you can't run 1920x1200 on the 20"/22"

Just a suggestion.......:)


~Ant
 
Last edited:
ok guys i've finished editing my form. we now have the following options.....

'1280 x 1024 HIGH', (this thread)
'1680 x 1050 HIGH',
'1920 x 1200 HIGH',
'1280 x 1024 VERY HIGH',
'1680 x 1050 VERY HIGH', (my exisiting thread)
'1920 x 1200 VERY HIGH'

obviously the very high is dx10 only.....

so get submitting. you can have one result of each type. if there is enough interest, i may start another thread to show the results which aren't covered by the existing threads. i don't want to hijack this one too much..... :p

http://www.marc2003.ukfsn.org/ocuk/submit.php?bench=crysis

do people really want AA/AF options added?
 
You're a credit to the forum and to the human race in general, marc :)

Thanks!


edit - I don't see the need for AA to be honest. Does anyone use it with Crysis? After all, Crysis has edge antialiasing built in by default already, and any AA you apply is the horribly expensive supersampling method...
 
i just tried 1920 x 1200 very high dx10 and was getting 11fps,but was very laggy indeed.
4850 overclocked to 700 and 1090.
 
marc2003 said:
ok guys i've finished editing my form. we now have the following options.....

'1280 x 1024 HIGH', (this thread)
'1680 x 1050 HIGH', 4xAA
'1920 x 1200 HIGH',
'1280 x 1024 VERY HIGH',
'1680 x 1050 VERY HIGH', (my exisiting thread)
'1920 x 1200 VERY HIGH'

obviously the very high is dx10 only.....
Click to expand...
Hi marc2003,

you make it looks easy, thanks very much again for your time, you are proving a most valuable contributer! :)

I think it would be great to get some AA results in there, after all isn't anti-aliasing not one of the most discussed techs on graphic cards? Even though I use a 1920x1200 monitor I wouldn't mine one bit if you scapped that catagory in favour of 1680x1050 4xAA HIGH, that would be a *most* useful resource and would really shed some light on how nVidia and ATI cards are doing in the high end.

This would a great ready reckoner for people that wanted to play this game (or WARHEAD or similar engine) and would reflect actual real world useage.

I played a lot of the game at high res using MEDIUM Quality settings and thought it looked good but to use HIGH quality I had to drop the res down to 1280x1024. I have no idea of what VERY HIGH settings looks like but isn't that a VISTA only option? (apart from hack?).

I think what I'm trying to say is I'd love to see more variation using AntiAliasing instead of just variable resolutions. See what the other lads say but in the end its up to you, thanks! :cool:

1680x1050 4xAA please!!!!!!!!!!
 
Last edited:
you make it looks easy
Click to expand...

that's because it was. i've done all the hardwork already. i only literally have to type in each setting i want. the rest is taken care of automatically. :)

i spent a lot of time on this when i first set it up ironing out bugs and whatnot but maintaining it is easy. and it's much easier than trying to keep track of a thread updating scores manually like i know many people have done in the past. even now there are a few guys maintaining excel spreadsheets manually in the overcloking forums for cpu speeds/pi times etc......

anyway enough waffle. i've gone with your tips and removed the options for 1920x1200 for now. and just gone with 1680x1050 HIGH 4xAA......
 
You must log in or register to reply here.
Back
Top Bottom