HD 4890 can beat GTX 285 and GTX 275 on CRYSIS WARHEAD benchmarking,by me and Friends

[Gaul]

Gaul: I was agreeing with you : P

Never mind .....we all here to sharing...THAT's ALL !!

[RPGWiZaRD]

First of all, I'd like to mention the 190.38 has horrible performance for some setups in Crysis, it has to be a bug because I've seen reports up to 50% FPS drop. The NVIDIA drivers have been very bad lately, the last good driver IMO was 182.50 WHQL, I personally prefer 182.47 but they are quite similiar.

That being said here's my result but at 1680x1050 res because I'm using a CRT and the 1440x900 res isn't selectable and the display driver I'm using doesn't allow me to add custom resolutions.

C2D E8400 @ 500x8, 4.0GHz
GTX 260 SP216 55nm @ 734/1512/1188 MHz (core/shader/mem)
Xtreme-G 182.47 driver
2x1GB DDR2-1000 4-4-4-8
Windows 7 64bit RC1

1680x1050, AA4x, DX10, Enthusiast setting, map: frost, 3 loops

Min: 27.07
Max: 51.85
Avg: 40.22



So either your systems aren't configured properly or you're suffering a lot of performance due to a bugged nvidia driver. I beat all those results and at a higher res, I know my CPU is clocked at 4.0GHz and also a dual core but you already said CPU performance won't matter much for this game anyway.

[Gaul]

182.47 but they are quite similiar.

performance due to a bugged nvidia driver. I beat all those results and at a higher res, I know my CPU is clocked at 4.0GHz and also a dual core but you already said CPU performance won't matter much for this game anyway.

so let me say thanx to u from 3 person, me ( gaul ), bxxco, Oxxn

THANX ALOT !!!!

LETS US TRY 182.47 !!

[Splave]

how much ram in each system?

[RAID]

I posted a GTX 295 SLI vs HD 4870X2 Quadfire comparison on here and got criticized for several reasons. I'll never waste my time w/ a forum review again.

[blaxtr3m3]

I posted a GTX 295 SLI vs HD 4870X2 Quadfire comparison on here and got criticized for several reasons. I'll never waste my time w/ a forum review again.

Don't get beat down by a couple of people who criticize you !!

Remember always how many you helped and made them take a good choice by posting good reviews and personal experience

Keep positive

[railmeat]

@ ALL = i guess Q66 @ 3.8 GH,Q94 @ 3.8 GH, Q955 @ 3.6 GH HAS A smiliar Perform !!!

"Q66 @ 3.8 GH,Q94 @ 3.8 GH,HAS A smiliar Perform !!"

clock for clock the q94 is way faster i would imagine.....anyone wanna chime in here?

[fatguy1992]

"Q66 @ 3.8 GH,Q94 @ 3.8 GH,HAS A smiliar Perform !!"

clock for clock the q94 is way faster i would imagine.....anyone wanna chime in here?

I don't see how it would be, sure its the newer 45nm etc, however hasn't it got smaller cache (6MB vs 8MB)?

[Gaul]

"Q66 @ 3.8 GH,Q94 @ 3.8 GH,HAS A smiliar Perform !!"

clock for clock the q94 is way faster i would imagine.....anyone wanna chime in here?

One again ! , hope the last, SIMILIAR PERFORM ON CRYSIS BENCHMARKING FPS !!! not super pi/CPU score VANTAGE/CPU score 3D mark 2006 !

IF on super pi/CPU score VANTAGE/CPU score 3D mark 2006 --------> YES it is BIG different, BUT not FPS on CRYSIS WARHEAD BENCHMARKING/GAMES

[blaxtr3m3]

Crysis is one of the most cpu demanding games IMHO. The better the CPU the brtter the gameplay !

[Demo]

well another nice statement !
the trully is 4890 very easy to touch 950/1150 even for daily !!! let me show u the max temp for 1000/1150 !! i think still comfort for my ear and belonging of the GPU !!

21mins of running Furmark is no where near the indication on how high the temps go, 3hrs atleast is the way to go

[Gaul]

21mins of running Furmark is no where near the indication on how high the temps go, 3hrs atleast is the way to go

i think 1 hours preety cool !



@ braxtreme = must see this again !

i agree 4 u'r statement if " c2q @ 3.8 GH better than c2d @ 3 GH " not in my case which q66 @ 3.8 GH with q9550 @ 3.6 GH, i ment q66 @ 3.8 gh not bottleneck to 4890 .

[blaxtr3m3]

That's 1024X768 m8 !!! with an x38 chipset and really old drivers !!!
also the faster the CPU the better the scaling for Multiple videocard SLi systems.
and you should not base all your conclusions on a single review !!

[Gaul]

and you should not base all your conclusions on a single review !!

ok i'll compare by my self later, now plz show me link which another review ...i would thanx for that !!

[mk-ultra]

1fps average more, big freaking deal, well within your margin of error due to different systems used

[Gaul]

am I have to bench again with q66@ 4.05 GH + 4890 @ 1010/1150 ?

[SamHughe]

should have just used one system to do the testing on btw.. alowise its hard to say its concrete proof :/

when testing try and use the same equipement and swap the video card.

9400 , 9550 , 6600 iam sure there is more differences than the CPU.

Exactly, when doing a test like this its pretty mandatory to use the same computer. Its like a crime scene investigation. You don't go to a crime scene before gathering evidence and start moving furniture around.

Let alone horrible language. I don't mind though.

As it has been already pointed out, comparisons like this are useless. This is because the fact that there is no way to know which card is fastest because the speed difference is caused by the CPU, RAM and OS used, maybe even more than by the actual GPU.

I did a comparison on a same system a while ago HERE

Here's my results based on Crysis Demo (1920x1200 all maxed):



And here's what I got from the Warhead Benchmark:

Single:



Crossfire:

[mk-ultra]

unless you use the same mobo/ram/hdd/fresh win install/ i don't see 1fps saying anything

[Eastcoasthandle]

unless you use the same mobo/ram/hdd/fresh win install/ i don't see 1fps saying anything

So where are your benchmarks that relate to the discussion of the OP?

[Gaul]

@ ALL = SAMHUGHE SPEC really Killer REALLY HUGE SYSTEM ! THANX FOR YOUR COMPARE !!!

CPU = W 3520 @ 2.6 GH
MEM = 12 GB
4890 @ 900/100 I'll COmpare !!!!!!

REMEMBER SAMHUGHE CRYSIS SETTING !!, sam i'll FIGHT u WITH only Q 6600 @ 3.4 GH !!!!
THIS WHAT HIS GOT !!!



MIN : 21.71
MAX : 39.09
AVG : 31.48

" AND MY OLD SYSTEM WILL TRY TO PROVE, WHAT I SAY ! "

CPU = Q 66 @ 3.4 GH
RAM = 4 GB FSB 472, cl 5/5/5/15
4890 @ 900/1000 ---. same with SAMHUGE

LET's SEE THE FINAL RESULT ! expecially for who said more faster CPU, will perform more !!!



MIN : 22.95
MAX : 40.18
AVG : 31.52

WELL, I HAVE SIMILIAR FPS, even little more than SAMHUGHE !
REMEMBER " SAME VGA 4890 @ 900/1000, SAME CATALIYST 9.7, and "little" different of CPU w3520 vs q6600, amd another "little" different RAM 12 GB vs 4 GB

@ blaxtr3m3,railmeat = i'll told u BEFORE !!

[RPGWiZaRD]

That's cool, I get about same performance as HD4890 crossfire with a single GTX 260 SP216.

EDIT: Ooops, he used "ambush" demo, not frost.

[blaxtr3m3]

I'm not here to flame your thread m8

I'm just stating what I know based on personal experience and reviews I read long ago (I will look them up for you).

To test my theory on your system, Just run the benchmark @ stock clock then overclocked, Then compare it to a core i7 system.

You should know that its not a linear change, Rather exponential.

that's y u had only 1Fps between the 2 clocks you used.

Also as the number of video cards increases the better the cards scale.

[Gaul]

I'm not here to flame your thread m8 :shakes

sory who is m8 ??? me ?

[kuchaiii]

sory who is m8 ??? me ?

m8 = mate

Yeah it's you... In your local forum used to say 'bro'

[blaxtr3m3]

sory who is m8 ??? me ?

yes