FX5700

[zeberdee]

Hi all.
      Does anyone know aanything about the Nvidia  
MSI Gforce FX5700 VTD256
graphics card, is it better than the "LE" or standard 5700? Many thanks Chris.

[Gixer]

Here is a VGA card chart



The LE isn't on it but I would say its a 'Light' edition which wont be anywhere near as good as a standard 5700.  Personally outta last years cards I would go the ATI route as they generally outperformed the equivilent Nvidia card and had better image quality.

[GreG]

What processor and Ram was used for those benchmarks?

[Gixer]

I dunno but it doesn't matter.  The same rig woulda been used with different cards so the results are comparible to each other.   The charts aim is to show the difference a g/card has only.

If I remeber right what ever it was run on the settings were not all maxxed though.

[congo]

Forget about the 5700 series, even the cheapest 5900 will outperform any of them at less cost.

You are better off with a 5900XT even, I have one and it overclocks easy to near ultra standard.

A better buy in the budget range is now the 9800pro, which can be found for around $150 US. Next step up I would choose a cheap 6800 card. That will be a huge leap again in performance.

That benchmark is good because it shows the results with the sim "working" hard on the hardware. It gives us an idea how the various cards perforn under load.

[4_Series_Scania]

I dunno but it doesn't matter.  The same rig woulda been used with different cards so the results are comparible to each other.   The charts aim is to show the difference a g/card has only.

If I remeber right what ever it was run on the settings were not all maxxed though.

Quite incorrect, it matters greatly!

For example, GeForce 6800's are CPU limited, i.e. they won't perform to their optimum with say a 2.0ghz p4 compared with a 3.2ghz p4, equally RAM types and settings can make a PC fly or equally, if the wrong type, crawl.!

To obtain the resulys posted in Gixers favorite piccy, Toms Hardware used this spec PC...

AGP Test System

CPU Intel Pentium 4 3.2 GHz
FSB 200MHz
Motherboard Intel D876 PBZ
Memory 4x Corsair CMX256A-3200LL DDR, 1024MB
HDD Seagate Barracuda 7200.7 120GB S-ATA (8MB)
DVD Hitachi GD-7000
LAN Netgear FA-312
Power Supply Antec True Control 550W
Drivers & Configuration
Graphics ATI Catalyst v4.9
NVIDIA v65.75
S3 v15.10.11.c
XGI Reactor v1.05
Chipset Intel Inf. Update
OS Windows XP Prof. SP1a
DirectX DirectX 9.0c

The page is here....  http://graphics.tomshardware.com/graphic/20041004/vga_charts-09.html
Hardly an average PC by anybodys standards! Try to get the same results with an XP 1800 or p4 2.4  ;)

[Gixer]

Ya ok, but with that spec it wouldn't matter so much.  The charts aim is to show how each type of g/card reponds to FS2004!

Yes they will not perform as well in a lesser spec base system but the cards further up the list would still be better than the ones lower down!

[4_Series_Scania]

The cpu limitations of the high end cards will mean they actually could under perform when matched to older slower kit, the same sort of kit that for example would return excellent fps with a ti 4600 or 9700pro.

I've tried a 6800GT in my #2 machine, a 1.8p4, it was certainly better than the FX5200 thats in it now, but, it scored roughly12000 marks in 3D Mark 2001 SE! a near identical score to my current PC and my FX5600 - a crap card by anybodys standards  :-[ - I tried the same 6800 GT in my #1 machine and it scored 20,000+ out of the box!  8) my point being, my other pc couldnt drive the 6800GT properly but my considerably quicker #1 machine (which I run @ 3.0ghz....) obviously could.

The 6800 GT ran approx 20% slower when I undid my overclock, returning my cpu to 2.6ghz.
To my mind,evidence of the dreaded "cpu limiting"

[congo]

I wouldn't be too conclusive on those results Paul.

I would run more tests (3dmark'03) and actually use the simming software to determine whether or not a video card is going to be held back.

As an example, my cousin just bought a 9800pro and ran 3dmark'03 on his 1.2ghz athlon machine. The darn thing nearly scored what my rig did, which surprised us both as you can imagine. That video card put him straight into the modern gaming world.

The only time a video card is held back by the CPU is when the CPU is struggling with the software, (which is a lot in many modern games!).

The video card is quite happy to run to it's potential, whether the CPU is coping or not. The overloaded cpu will become apparent by reduced game performance.

[4_Series_Scania]

Hmm, I see your point congo!

I suppose something else must have been the culprit with my #2 machine..... 12,000 was a bit naff!  ::)

[Gixer]

Ooooer 12000 with a 6800GT? somthin very wrong.  On my old K7n2 Delta with 333FSB and an XP2800 and a GF4 Ti4600 I was scoring over 13000 lol.

Maybe it was just a driver conflict.  These newer component seem to be extremely fussy on drivers and bits of old drivers being left on your PC etc.

[Delta_]

Ooooer 12000 with a 6800GT? somthin very wrong.  On my old K7n2 Delta with 333FSB and an XP2800 and a GF4 Ti4600 I was scoring over 13000 lol.

You get that score with the Ti4600 in 3Dmark01, the 6800GT gets it in 3Dmark03.

[Gixer]

You get that score with the Ti4600 in 3Dmark01, the 6800GT gets it in 3Dmark03.

Lol I know that and its more like around 11,000.  But we are talking 3dMark01 which tests your whole system rather than just your G/card.  Thats y we sayin somthin is a bit weird here!  My Ultra scores me over 13000 in 3dMark03 and my current setup scores just under 24000 in 3dMark01.

Saying that though 4_Series it may be about right. Out of interest how much higher did it score than the FX5200?  It should be quite a bit.  Though saying that When I played around changing my FSB I saw vast improvements in 3dMark01 scores thus saying it is probably not a totally true representation of how good your g/card is!  3dMark03 is better for that and we all know the 6800GT will oblitorate the FX5200 in that hehehe.  

[Gary R.]

Just remember. We humans can still only percieve 24fps. Its an anatomical fact. (Cats can percieve up to 40 fps).

[Gixer]

Yup totally but I have future proofed myself and can run games with max settings all the eye candy and loadsa AA and AF.  I could never do that with an FX5800 specially on the newer games!  Plus I had some spare cash and have never had a top line G/card and I am extremely glad I have it now as my games are just awsome.

I dont believe its been prooved a human cant see more than 24fps either!  every one sees stuff differently, just ask on here! I for one can see the difference in smoothness of the game from 20-35 FPS.