Forum Settings
       
This Forum is Read Only

Benchmark OdditiesFollow

#1 Jul 02 2010 at 8:16 AM Rating: Good
***
3,105 posts
So I tried the benchmark on my 3 computers...

Computer #1, oldest, wouldn't run the benchmark at all, it crashes every time you try. I pretty much predicted this; it is an older WinXP machine that barely handles 60 FPS in WoW at lowest settings.

Computer #2, in the "Spare Room", scored a 2100. It has:

Athlon64 x2 5400+ Brisbane 2.8Ghz
Northbridge: AMD 780V
RAM: 4GB DDR2 1066 SDRAM
GPU: Radeon 4830HD 512MB

Computer #3, in My Room, scored 1820-1870 (ran it 3 times). It has:

AMD Phenom 9750 2.4Ghz Quad-Core
Northbridge: AMD 790FX
GPU: Radeon HD 4670 512MB
RAM: 8GB DDR2 1066

So my question is, why is computer #2 scoring so much higher than computer #3? I changed the desktop resolutions (why the Benchmark App doesn't Full-screen is beyond me, I bet I'd get 500+ score higher if it were fullscreen). Is it the CPU, or the Graphics Card maybe? I'm trying to figure out which is the weak link in Computer #3...

What confuses me even more is that Computer #3 gets higher FPS in WoW at the same settings than Computer #2... but yet Computer #2 scores higher on the benchmark... but Computer #3 loads it up faster (the "loading time" at the end I mean).

BTW: Both are running the same version of Win7.

Edited, Jul 2nd 2010 10:18am by Lyrailis
#2 Jul 02 2010 at 8:19 AM Rating: Good
Sage
**
743 posts
I believe that the benchmark relies heavily on the GPU. Your 4830 has more horsepower than the 4670.
____________________________
I think you've been smoking the Moko...
http://na.finalfantasyxiv.com/lodestone/character/350413/
http://na.finalfantasyxiv.com/lodestone/character/1628942/
http://www.nerdist.com/
Angus of Cerberus (retired)
#3 Jul 02 2010 at 8:23 AM Rating: Good
***
3,105 posts
AngusX wrote:
I believe that the benchmark relies heavily on the GPU. Your 4830 has more horsepower than the 4670.


Hmm, there'd be an easy way to test it -- just swap the cards. lol.
#4 Jul 02 2010 at 8:24 AM Rating: Decent
Posting your score is pretty meaningless unless you also include what resolution setting you ran the benchmark at. Different resolutions will bottleneck at different components.
#5 Jul 02 2010 at 8:26 AM Rating: Good
***
3,105 posts
The One and Only Aurelius wrote:
Posting your score is pretty meaningless unless you also include what resolution setting you ran the benchmark at. Different resolutions will bottleneck at different components.


Oh, d'oh.

Low Res of course.

Desktop Resolution, I forget what the Spare Room Computer had, but this one had 1440x900 and then tried it again at 1280x800.

Same score (within 30) both times, so I'm convinced that desktop resolution doesn't have an impact within reason.
#6 Jul 02 2010 at 8:29 AM Rating: Good
Scholar
***
1,707 posts
My guess is that both have plenty enough CPU power to pump data to the Video Card fast enough not to cause a bottleneck. From there it's all on the video cards, and clearly Machine #2 has the more powerful card. So yes, you are right that Video Card means much more than all the other stuff, assuming "all the other stuff" isn't so bad that it's causing a bottle neck.

#7 Jul 02 2010 at 8:49 AM Rating: Good
***
3,105 posts
Update:

Switched the cards, Computer #3 scored a whopping 2705.

Jeez.

Yeah, okay, it was the video card for sure.

Model Numbers can be so misleading, both are PCI-E x16 512MB cards, and one is an "HD 4830" and the other is an "HD 4670" so they *sound* pretty close... meh lol.

Thanks for the suggestion!

Edited, Jul 2nd 2010 10:50am by Lyrailis
#8 Jul 02 2010 at 10:10 AM Rating: Good
Sage
**
743 posts
I've got a 4650 that I've tried in 2 computers. Both have 4 GB ram with W7-64bit. One has a e2160 (1.8 GHZ) cpu, the other an OC'ed i3 (3.7 GHZ). Scores with the GPU OC'ed to max in ATI Overdrive: 835 in the first, 885 in the second. IMO, video card is a major factor for FFXIV. I'm wondering if better CPU's will have the advantage in crowded areas.
____________________________
I think you've been smoking the Moko...
http://na.finalfantasyxiv.com/lodestone/character/350413/
http://na.finalfantasyxiv.com/lodestone/character/1628942/
http://www.nerdist.com/
Angus of Cerberus (retired)
#9 Jul 02 2010 at 11:18 AM Rating: Good
***
3,105 posts
AngusX wrote:
I've got a 4650 that I've tried in 2 computers. Both have 4 GB ram with W7-64bit. One has a e2160 (1.8 GHZ) cpu, the other an OC'ed i3 (3.7 GHZ). Scores with the GPU OC'ed to max in ATI Overdrive: 835 in the first, 885 in the second. IMO, video card is a major factor for FFXIV. I'm wondering if better CPU's will have the advantage in crowded areas.


CPU is also a part of it too.

There was a 4GB difference in RAM, so what I did, was I took some of Computer #3's RAM and put it in Computer #2 so that Computer #2 had 8GB of RAM (and yes, before anyone asks, I _am_ running 64-bit Win7...).

Computer #2 then scored 1880 instead of 1820.

That's not much of an increase.

So, Computer #3 with the better card will score 2700 while Computer #2 with the same card will score 1880, both with 8GB of RAM and identical Windows7 64-bit OS's.

Identical A/V software, the only difference Software-wise, is Computer #3 is running Daemon Tools in the background, whereas Computer #2 is not. That should actually _help_ Computer #2's case.

The only remaining differences are the Motherboard and CPU; Computer #3 has a Quad-Core 2.4Ghz while Computer #2 has a Dual-Core 2.8Ghz.

The Northbridge Chips are also slightly different, but I'm wondering how much of a difference that would actually make. If it weren't for the extreme annoyance of having to discombobulate the CPU Fan/Heatsink/etc, I'd try dropping my Quad-core in the other computer and see what happens... but I'm not feeling _that_ energetic....
This forum is read only
This Forum is Read Only!
Recent Visitors: 24 All times are in CST
Anonymous Guests (24)