Benchmark results for NVIDIA GeForce FX Go5200 (mew905)

[mew905]

Dell Inspiron 5150

BIOS A38 - HT On - Speedstep Off

Intel Pentium 4 M 2.8 GHz @ 908MHz

Mobo Driver version 5.00.1012, A06

512MB DDR PC2700 RAM

GeForce FX Go5200 64MB (196/392 stock)

Forceware v67.42, A08

All are most recent drivers to date (May 11, 06)

I've decided to benchmark this video card at 4 clock speeds. The first is the stock speeds, the second is what seems to be what most gamers prefer to clock at (between 225 and 240 clock, 425 and 450 mem), the third is my clock settings at which I play games at and is most stable, and finally the 4th clock is pretty much just for kicks to see the results of an even higher clock. I did try 255/495 but it was too unstable to benchmark at all. Keep in mind that not all the time are you going to play games at the default settings (unless of course it has an automatic, hardware-scanning graphics code), which is why I chose to run the benchmarks at the low settings.

The percentages after the score/fps is the performance boost you can expect to see over the stock clock speeds.

Doom 3 - Low Settings -NoAA NoAF - 640x480 Timedemo 1

196/392 (core/mem) - Avg. FPS = 16.0 +0%

225/450 (Core/mem) - Avg. FPS = 18.2 +13.75%

244/477 (core/mem) - Avg. FPS = 19.9 +24.375%

245/490 (core/mem) Avg. FPS = 19.7 +23.125% (1.015% slower than 244/477)

NOTE: Seems the sweetspot for my particular card is 244/477 as the Inspiron dropped to 3 FPS twice for about 3 seconds during the last test. The first two tests went without a hitch. Please also note that I usually get about 15-30 FPS ingame, which for doom 3 is playable.

3DMark03 - NoAA NoAF - 640x480 - 4 Game benchmarks

196/392 (core/mem) - 1974 3DMarks +0%

225/450 (core/mem) - 2194 3DMarks +11.145%

244/477 (core/mem) - 2301 3DMarks +16.565%

UPDATE: Using a IRQ Priority registry tweak, I managed to gain a bit more performance! Not much, but every bit helps.

244/477 (core/mem) - 2322 3DMarks +17.629%

245/490 (core/mem) - 2331 3DMarks +18.085% (1.304% faster than 244/477)

3DMark03 - NoAA NoAF - 1024x768 - 4 Game Benchmarks

244/477 (core/mem) - 1466 3DMarks

NOTE: Once again that sweetspot is there. 245/490 may have scored higher, but it also faced some noticable visual artifacts. They were more noticable during the last of the 4 tests.

The rest of the tests are run at 244/477, as the above tests can be used as a comparison. If anyone objects to this, I will test at the other clock speeds as well.

Alien Versus Predator: Primal Hunt - Max settings

Frames: 8652 - Time: 298781ms - Avg: 28.957 - Min: 15 - Max: 60

This is a comparison computer:

3DMark same settings as above tests

Intel Celeron 2.3GHz @ 2.38 GHz

512MB PC2700 DDR RAM

BFG GeForce 6600GT OC (525/1080)

3DMarks = 9734

At 1280x1024 (I was trying to slow it down to the same score as the FX Go5200) w/ 4xAA 4xAF, it still scored a little over 3000 3DMarks

If anyone wants other games benchmarked, I do have access to alot of them. Starcraft, Sim City 4, Sims 2, C&C Generals, Need for Speed U, U2, and Most Wanted, Counter Strike, I can get Half Life 2... Just post here.

Edited June 13, 2006 by mew905

[sn0wl]

Sound very interesting. I would like to pin this topic if you are willing to continue benchmarking, are you?

[mew905]

yes, I am willing to continue benchmarking. If someone wants a certain game to bench so they can see what they can expect with a system similar to mine.

[Guest Josh]

Was just trying to get my graphics card to work better. Read the overclocking thing and though i'd whack mine up to the settings you use. However when going into the overclocking tab, my geforce 5200 go seems to have default clock settings at 230 core / 513 memory.

The core is lower than yours but the memory seems to be clocked way too fast. These are the default settings the laptop came with.

Just wondering if anyone knows which is really more important - memory of core gpu speed?

Cheers,

Josh

[mew905]

they're both just as important. Data corrupts from being too fast. if those are stock speeds and you want to overclock, divide the memory clock in half and clock the core up to what the result was. (about 256MHz) My stock speeds were already mentioned, and this is as fast as my card will go without extra voltage or artifacts.

When either is clocked too high, artifacts will appear, I like to think it's because the core is too fast for the memory, the memory gets confused, and sends out the wrong data, and vise versa when mem is clocked faster. The general rule of thumb is to keep memory at 2x speed of the core (it's DDR so it's showing you the effective speed, it's actually only running at half that) Obviously that doesnt have to be done, as mine is stable at 244/477. If those are your stock speeds, you should easily be able to hit overclocked FX5200 speeds (non-go card). With my FX5200, I can grab 315/550 (if I remember correctly)

[daonesteven]

meo, if you put it at that overclocked clock speed, do you leave it at that speed all the time? or do you only overclock it to those settings for games? and if you say yes to the latter, do you manually change those memory settings every time before you start a game? i've been trying to find a way to have like clock setting profiles that i can easily switch to with a click of a button or something ><

also...how come you dont use sudlow's 5150modified 67.66 driver? did you find that 67.42 was better?

Edited May 21, 2006 by daonesteven

[mew905]

I dont really find any forcewares are better than others. If you read the 5200 benchmark that lists the drivers and 3dMark scores, you'll see each driver is better in different areas. I play a wide arrange of games so modifying the driver to suit 1 game doesnt really work out for me.

As for the clock settings, I leave it there constantly, the only time the video card really heats up is during games, overclocking just for games just seems like an inconvenience. Its like a CPU, the only time it starts heating up is when it is at a higher load. of course I'm talking the CPU idle temp is the lowest in that instance. As for the clock speeds at the click of a button, I dont know how you would do that, I could research and build you a program that will either increase clocks at the click of a button, or increase clocks at the start of a game. Actually, I may even make myself one.... Powermizer doesnt do much... just slows things down until unplayable.

[mew905]

Added a couple updates, just minor. Included a comparison computer, a new benchmark for 3DMark03, and a benchmark for AVP2 Primal Hunt using FRAPS.

[Guest Guest]

BIOS A38 - HT On - Speedstep Off

Intel Pentium 4 M 2.8 GHz @ 908MHz

HT on?

does it only work with the pentium mobile-version?

i have an instpiron 5150 too, but a pentium 4 processor at 3060 mhz.

you know if i can turn on HT here?

anyway, nice results.

[Ultima]

The older Inspiron 5150s didn't have HT (mine is one of the older ones).

[mew905]

I didnt know if older I5150's didnt have hyperthreading. Personally I find that it has almost no effect on performance. Mine does have HTT (Hyperthreading Technology) however.

[Sevyndaze]

Hi, New here, but I'm running almost your identical setup, except my card is only 32MB. I'm running the forceware 84.25 driver with a modified INF to fix the powermiser settings. I followed the overclocking guide exactly, but I had a small question about the 3Dmark software. The first game test runs very vast, though I don't know the exact number, it got up into the hundred-some range on several occasions, and never drops below 60. However, every subsequent test runs horribly, averaging more in the area of 10-15 frames per second. I am currently running it overclocked at 243/486, yielding a 3Dmark score of 1126....the best of about a dozen configurations i tried. Anyone have an idea what could cause such a drastic difference between the first game and the rest? Any help would be much appreciated.

[mew905]

first test isn't nearly as graphic intensive. my card reaches a bare minimum of 85 fps, and goes upwards of 160. the second test (trolls lair) runs an average of about 12 fps, the next runs an average of about 7, and the final test runs an average of maybe 9-12 fps. the extra 32mb of memory will help for the more intense games, but its not much.

I use the latest dell drivers which are 67.42, which works fine for power mizer settings for me. later modded inf drivers the powermizer feature is locked. your getting about average performance and remember, my tests were at 640x480.

[Sevyndaze]

Ok, I tried running some games today, namely Half Life 2 and Need For Speed: Most Wanted. Both run very well at 1024x768 with good detail settings, except every half-minute or so it seems like it skips or lags for a second, and then normal play resumes. I'm such a noob to this computer/videocard tuning stuff...any suggestions?

[Guest Guest]

probably a heat issue. my laptop does it as well, randomly, but occurs more often when the fan is covered.

[Guest mew905]

likely a heat issue yes, but I've had the same problem when I clock the memory too high without the core bein up there too. It lags, sticks for about 3-5 seconds, then resumes, sometimes you'll get artifacts. My laptop doesnt overheat in those conditions (especially with winter comin up here quick).

I understand I'm boosting up a thread over a year old, sorry if that's against the rules.

Unfortunate though that my laptop is so outdated now I cant run the newest games (they all require a DX9.0c video card bare minimum now). Of course there are a few that I can, but the latest game I've been able to run on it was NFS Carbon at minimum settings. I usually get around 45fps there, 60fps on the TV screen for some reason (the settings didnt change, the TV is capable of 1024x768 and actually handles high-def very well! it's a CRT TV from 2000).

so looks like it's stuck to games around a year old... kinda sucks... reminds me of the old laptops like the P2 with onboard, it'll be in that class fairly soon... makes me sad because I paid like $3000 for it (3 year complete-care, all the goodies except built-in wireless). Thats one thing that completely sucks about laptops, almost no upgrade possibilities... Ah well, I got my new gaming rig.

[Guest Kim]

If you are still benchmarking, can you please let us know if Sims 3 will work. These are my specs:

Time of this report: 1/17/2009, 21:50:17

Machine name: DELL-C82A271F9F

Operating System: Windows XP Professional (5.1, Build 2600) Service Pack 3 (2600.xpsp_sp3_gdr.080814-1236)

Language: English (Regional Setting: English)

System Manufacturer: Dell Computer Corporation

System Model: Inspiron 5160

BIOS: Phoenix ROM BIOS PLUS Version 1.10 A08

Processor: Mobile Intel® Pentium® 4 CPU 3.06GHz (2 CPUs)

Memory: 1280MB RAM

Page File: 734MB used, 2899MB available

Windows Dir: C:\WINDOWS

DirectX Version: DirectX 9.0c (4.09.0000.0904)

DX Setup Parameters: Not found

DxDiag Version: 5.03.2600.5512 32bit Unicode

------------

DxDiag Notes

------------

DirectX Files Tab: No problems found.

Display Tab 1: No problems found.

Sound Tab 1: No problems found.

Music Tab: No problems found.

Input Tab: No problems found.

Network Tab: No problems found.

--------------------

DirectX Debug Levels

--------------------

Direct3D: 0/4 (n/a)

DirectDraw: 0/4 (retail)

DirectInput: 0/5 (n/a)

DirectMusic: 0/5 (n/a)

DirectPlay: 0/9 (retail)

DirectSound: 0/5 (retail)

DirectShow: 0/6 (retail)

---------------

Display Devices

---------------

Card name: NVIDIA GeForce FX Go5200

Manufacturer: NVIDIA

Chip type: GeForce FX Go5200

DAC type: Integrated RAMDAC

Device Key: Enum\PCI\VEN_10DE&DEV_0324&SUBSYS_01961028&REV_A1

Display Memory: 64.0 MB

Current Mode: 1024 x 768 (32 bit) (60Hz)

Monitor: Default Monitor

Monitor Max Res:

Driver Name: nv4_disp.dll

Driver Version: 6.14.0010.6610 (English)

DDI Version: 9 (or higher)

Driver Attributes: Final Retail

Driver Date/Size: 8/19/2004 00:44:00, 3763072 bytes

WHQL Logo'd: Yes

WHQL Date Stamp: n/a

VDD: n/a

Mini VDD: nv4_mini.sys

Mini VDD Date: 8/19/2004 00:44:00, 2973568 bytes

Device Identifier: {D7B71E3E-4064-11CF-4A6A-9C2100C2CB35}