Share
| Welcome to Evolved. We hope you enjoy your visit. You're currently viewing our forum as a guest. This means you are limited to certain areas of the board and there are some features you can't use. If you join our community, you'll be able to access member-only sections, and use many member-only features such as customizing your profile, sending personal messages, and voting in polls. Registration is simple, fast, and completely free. Join our community! If you're already a member please log in to your account to access all of our features: |
| Benchmarking | |
|---|---|
| Tweet Topic Started: Oct 26 2005, 10:30 AM (172 Views) | |
| Maplassie GTR | Oct 26 2005, 10:30 AM Post #1 |
|
MR GTR
![]()
|
Do you trust it, do you beleive in it? Do you beleive that bencharks accuratly sums up your pc performance? I think not. While Benchamrking gives you a rough estimate of the capabilities of your machine, I never ever ever trust benchmarks. Yes I do run them, but only for fun, and I never care what the results are! 1.Let me tell you some things I have encountered. about a month a nd a half ago, I ran 3dmark 2005 3 times in a row. First up I got a score of 5500, wich was an unbeleivable score for a 128mb non ultra non gt 6800 card. I even posted you the piccie as proof. the second run produced a score of 3300. Yes, 2000 points drop without changing anything, the third run produced a score of 4600. And that's 3 times consecutavly. I have not run a 3dmark 05 bench since. 2. A while back, before I started posting on forums, I had an AMD XP 3000 Cpu, with 1gig samsung ram and a FX5950 Ultra VGA card. My friend had an intel 2.8 P4 with 1gig cheapass ram, and a Fx5600 card. IN EVERY EVERY EVERY EVERY benchmark, Codecreatures,Aquamark,3dmark 01,03,05 and even SISof sandra and PCmark, his pc, VGA,Ram and harddrive outperformed mine by a whole bunch. every single time in every single way. no matter what we did, or how we did it, all benchies left his machine the victor. Sound impossible right, but it's true. BUUUUUT. When it came to gaming, his mahcine could not touch mine. UT2003,UT2004, Farcry (when it just came out),deus ex2, any game, my machine outperformed his by a whole lot. I could run much higher settings, much smoother, faster loading times and all. Now is that weird or what? Hence, I dont trust benchmarking at all. |
| |
![]() |
|
| VoodooprophetII | Oct 26 2005, 10:43 AM Post #2 |
|
PC HATER
![]()
|
I'm sceptic of some benchmarks.......I do believe some, like super-pi and I do trust Aquamark3 to a certain extend because I can actually see my scores improve by adding better memory, CPU and GFX...and I know how to fine tune my system for Aquamark3. For me it's the quickest way to see if I've got a problem or not with my system. I've never had the random score changes...and yup 5500 on 3D Mark05 for a 6800nu is way high.... As for 3D Mark in general.....3D Mark 01 is synthetic, the same with SandraSoft. mean it's more a guesstimate than actual performance......like looking at clock speed of the cpu's. As for 3D Mark03 and 05...yup there it's more GFX than anything else, but the anything else will have some effect on the score. I don't like 3D Mark03 or 05 because between ATI and Nvidia....I feel both you'll find at some stage cuddling up to futuremark..... |
|
<a href="http://profile.mygamercard.net/SARELSLIPSTREAM"> <img src="http://card.mygamercard.net/gelsig/teal/SARELSLIPSTREAM.png" border=0> </a> | |
![]() |
|
| VoodooprophetII | Oct 26 2005, 10:43 AM Post #3 |
|
PC HATER
![]()
|
:banana: Dubble post |
|
<a href="http://profile.mygamercard.net/SARELSLIPSTREAM"> <img src="http://card.mygamercard.net/gelsig/teal/SARELSLIPSTREAM.png" border=0> </a> | |
![]() |
|
| someoneelse | Oct 26 2005, 10:49 AM Post #4 |
|
Through the looking glass
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
I do agree with both maplassie and voodoo. Benchmarks (especialyl 3dmark) don't reflect the true performance of a machine, but like voodoo said, it can give a clear indication of how well the machine might perform. Maplassie - I think your FX5950 got a lower score because it utilises different shader models, which weren't properly supported back then. But like 3dmark05 - you have to have graphical power to increase your score. It doesn't give an exact result - but is a good method of seeing if your system has a bottleneck of some sort. There's no ways you'll reach 7000 3dmarks in 05 by accident with an older card. |
![]() |
|
| Maplassie GTR | Oct 26 2005, 11:03 AM Post #5 |
|
MR GTR
![]()
|
dude, read my post again. My Friend had a FX5600 256mb 128bit wich cost R700 I ahd a FX5950 Ultra, King of the hill before the 6800 came out. and he outbenched me. So that rules out old technology, cause he had an older card than me, and much slower, and only 128bit. BUt I do understand what you mean though! |
| |
![]() |
|
| VoodooprophetII | Oct 26 2005, 11:29 AM Post #6 |
|
PC HATER
![]()
|
Hey Maps....I remember a whole story about drivers Nvidia brought out to increase benchmark performance....the whole detonator story, and later if I'm not mistaken, I think ATI had the same thing as well, which I would say is gonna include this new Catalyst 5.11 patch for increasing Doom3 (opengl performance). So if anybody's to blame for inaccurate bench results, look at Futuremark's pockets...and ATI and Nvidia .....and many other developers prefering one card to another.... |
|
<a href="http://profile.mygamercard.net/SARELSLIPSTREAM"> <img src="http://card.mygamercard.net/gelsig/teal/SARELSLIPSTREAM.png" border=0> </a> | |
![]() |
|
| someoneelse | Oct 26 2005, 11:31 AM Post #7 |
|
Through the looking glass
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
Yup. It all comes down to gaming... And I can tell you one thing: My X850 totally creams my 6800GT. The frame rates are fooking amazingly jaw dropping even at resolutions like 1600x1200. FOok mi I wish I could explain how cool it is... |
![]() |
|
| Maplassie GTR | Oct 26 2005, 11:34 AM Post #8 |
|
MR GTR
![]()
|
I know dude! I know.... I will have the same feeling soon......... |
| |
![]() |
|
| VoodooprophetII | Oct 31 2005, 09:27 AM Post #9 |
|
PC HATER
![]()
|
Okay....I had a funny thing happen to me......It is to an extend Benching related, but with a touch of something else. Inconsistencies when you benchmark will occur and sometimes it's totally reason to what might seem obvious.... Now some time back I got myself a DFI mobo and a PCIE gfx card, and with a shock I realised "oh bugger, my PSU has a 20pin male connector and the mobo a 24pin female connector.....funny enough the PSU plugged in without any problems...to the left of the socket, leaving 4pins open.....the 4pins was to com (or ground lines) and the other two pins is +3.3V and +5V which I though, tough sh1t, I mean my important line or rail is the +12V which has 30Amps and it's an Antec True550 PSU, so it should work. It does work and it works well. No consulting real overclockers, like Jasonhk, and Solo and some of the others, and the DFI street forum, they say, yup it will work, but they don't recommend it.... So I looked around for a better 24pin PSU....pricy they are.... While sitting with the 20pin Antec True550, I overclocked my GFX and CPU a bit....CPU form 2.4 to 2.8Ghz, with the Memory running DDR466 and I needed to up the vdimm to 3.4V for stability (drains tons of power), and the GFX I took to 570 and the memory at 1214...... I only benched Aquamark3 and I got 85000 from a 78900 stock. I thought hell, it's not too bad, but I expected more.....I mean with my 6800GT overclocked same CPU clocks I got 84700 but it's okay. And I needed to run the bench about 4 times before I get a run with no slowdown.....(and the slowdowns worried me) Friday I tested my sytem with a Thermaltake 680Watt PSU same overclock...exactly the same all over, and I ended up with 88890, which is almost 4000 points more.....and that was one run.....the first overclock run.....so ya, sometimes the factors influencing performance lies beyond the obvious..... I'll find the pics and post them for the proof. |
|
<a href="http://profile.mygamercard.net/SARELSLIPSTREAM"> <img src="http://card.mygamercard.net/gelsig/teal/SARELSLIPSTREAM.png" border=0> </a> | |
![]() |
|
| VoodooprophetII | Oct 31 2005, 09:31 AM Post #10 |
|
PC HATER
![]()
|
Here's four pics.....First two will be one bench result, and second two a second bench result, with the same overclock. The difference is not FSB, or memory timings or Quality settings....all of them were done with exactly the same settings. First bench the difference was that it was done with a Antec True 550Watt PSU and the GFX core is 10 Mhz more and memory 10Mhz less. Settings: CPU 280 x 10 Memory (166 div) to make it DDR466 at 2 2 2 5 T1 GFX 1st run 580 and 1200 GFX 2nd run 570 and 1214 HTT 280 x 3 Fist bench with the 20pin Antec True 550 You can see the clocks in this image: Posted Image This one you'll see the score and watch the min and max FPS as well as the GFX and CPU score Posted Image Second bench with the 680Watt PSU Clocks is the same: Posted Image Compair the CPU and GFX score to the 1st bench: Posted Image Now except for the slight GFX clock differences, the scores is actually much different to what I would have thought...... I did notice that when overclocking withthe Antec True550 PSU, it's okay, but you get these random slowdowns in Aquamark, and you have to run a few times to actually get a clean run. When I installed the 680Watt 24pin PSU my first run overclocked is the score I posted....no problems or slowdowns....shows you what a difference a PSU can make.....I can't really push it higher since all's on air, and CPU on stock HSF.. B) |
|
<a href="http://profile.mygamercard.net/SARELSLIPSTREAM"> <img src="http://card.mygamercard.net/gelsig/teal/SARELSLIPSTREAM.png" border=0> </a> | |
![]() |
|
| someoneelse | Oct 31 2005, 09:40 AM Post #11 |
|
Through the looking glass
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
Amazing. I'm using a 400 watt 20-pin PSU in my 24 pin mobo - you reckon I should be looking at a new PSU? Think 400 watt is enough for my system? |
![]() |
|
| VoodooprophetII | Oct 31 2005, 09:47 AM Post #12 |
|
PC HATER
![]()
|
Hey SOE.....I don't think it will have an huge affect if you running stock...my stock runs is exactly the same.....no difference, but as soon as you start overclocking, that's when you need more juice... As for the mobo you got, hell maybe they don't draw even close to the power the DFI mobo's draw...so maybe it's not a problem for you.....I mean your stock runs is in the same line as mine with 3Dmark05 and it will drain power for the bench, even at stock speeds..... Most ppl will not recommend a 20pin PSU on a 24pin mobo, but a lot of ppl do use them, and with no ill effects...so don't stress as yet. Check you +12V rail how many amps does it give you....that's the little sticker on the PSU... |
|
<a href="http://profile.mygamercard.net/SARELSLIPSTREAM"> <img src="http://card.mygamercard.net/gelsig/teal/SARELSLIPSTREAM.png" border=0> </a> | |
![]() |
|
| someoneelse | Oct 31 2005, 09:55 AM Post #13 |
|
Through the looking glass
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
I'll check... And I do overclock my X850 - I OC it to the same speeds as yours, memory 1200 and GPU 580 |
![]() |
|
| Maplassie GTR | Oct 31 2005, 10:37 AM Post #14 |
|
MR GTR
![]()
|
Bigger is better. Always. LOOL! |
| |
![]() |
|
| 1 user reading this topic (1 Guest and 0 Anonymous) | |
| « Previous Topic · Entertainment · Next Topic » |
| Theme: Zeta Original | Track Topic · E-mail Topic |
4:32 PM Jul 11
|










![]](http://z2.ifrm.com/static/1/pip_r.png)

4:32 PM Jul 11