Voodoo 5 vs Geforce Shootout

Falcon 4 Benchmarks & How To

Dancing with my Coppermine OC P-IIIe 550

Download Apps & Falcon 4 benchmark

HOME

Reviews

Voodoo5vsGeforce2

Flanker2 

P-IIIe 550 Overclock

V3-3500 vs V2 SLI

Falcon 4 Review

Flanker2-98 E3 Report

AMD K6-2 333

Sims/Tactics 
Benchmarks & Stories

Falcon 4 Benchmark

Flanker 2 & Su27 Base

Janes F-15 Base 

Real Life tm  "Tales"
Flanker 2 3dfx help

Fox2 Freeware
Falcon4 Graph tool

Downloads Page

Great Links
Military Information, Other Sims sites, Hardware sites

My Family Site

 

Sunday, June 10, 2007 09:22 PM

Benchmarking the Beast…or just how fast is it. 

Understand that I only changed the processor and this resulted in a very low cost of upgrade. Its important to keep the RCS (radar cross section for you non flight sim fanatics) of any upgrade below the detection level of the wifey…<VBG>. Processor and Slocket Adapter came to around 380.00 dollars. And as you will see for the purpose of improving my framerates in flight simulations it was money well spent. So here without further ado is my current machine configuration. (Yes the Voodoo 2’s are near the end of the line as my main 3d Cards these tests make that much obvious…<VBG>)

My machine consists of 
Abit BH-6 Revision 1.01 motherboard
128mb of Mushkin Samsung PC-100 –GH SDRAM
Windows 98se 
Tseng ET6000 2d Video Card
Twin Voodoo 2 12mb Cards clocked to 100mhz See notes 1
Three hard drives: See notes 2
            a.       IBM Deskstar 4.1 Gb Ultra DMA drive partitioned C:, F:, G:,
            b.       Quantum Fireball 12.9 Gb partitioned D:
            c.       Western Digital 1.6 Gb E:

  • Notes:

  • 1. The Voodoo 2 Cards are reaching the end of their usefulness. Only problem is I cannot make up my mind on what to get.

  • 2. The hard-drives have proven to be the most difficult problem for overclocking my particular setup

Results from 3D Mark 99 Max

 

Celeron 458

P-III 550e

P-III 682e

3D Mark Result

3,113

3,587

4,368

Synthetic CPU 3D speed

3,742

7,880

9,853

Game 1 Race

31.3

40.2

42.2

Game 1st Person

30.9

32.4

45.3

Looks to me like we are bumping up against the fill rate max for the Voodoo 2’s in the Race Bench. Note that all 3d Mark tests were run at 1,024 x 768.  

Results from Si Soft’s Sandra

 

Celeron 458

P-III 682e

CPU Benchmark

CPU: 1113 MIPS,
FPU: 550 MFLOPS

CPU: 1699 MIPS,
FPU: 853 MFLOPS

CPU Multi-Media Benchmark

CPU: Integer MMX: 1041, 
Floating-Point FPU: 605 it/s

CPU: Integer MMX: 1517, 
Floating-Point FPU: 921 it/s

Memory Benchmark

CPU/Mem: 219 MB/s,
FPU/Mem: 200 MB/s

CPU/Mem: 308 MB/s,
FPU/Mem: 286 MB/s

Game Benchmarks
This is the beginning of the pure game benchmarks. If the chip doesn’t rock here then all the speed in the world in synthetic benchmarks doesn't do any good.  Theory is good and valuable but you must confirm that its correct with benchmarks in the apps you use.

Results from Falcon 4 Timedemo 
This is a very reliable way to test your entire system. Because I have been using this time demo for such a long time I am able to get within a .5 difference on an unchanged system. The repeatability of the time demo is crucial for a worthwhile test.  

Dont forget to visit the Falcon 4 Benchmark page

 All settings on Max Detail

Falcon 1.08Ia 1024x768 
458mhz Celery

Falcon 1.08Ia 1024x768 
P-III 682e mhz

Average Max Framerate

16.9

26.4

Average Min Framerate

9.6

15.5

Average Framerate

13.9

21.7

As you can see with this probably the most important benchmark for me that indeed the upgrade has improved matters significantly.

Results from Flanker 2.01 benchmark mission
This was the sim that got me thinking that maybe it was time to move on from my Celeron 458 combo.  I put together a mission that stressed the system extremely hard and was repeatable. I am confident that this Benchmark has no more than a 1-frame error either way. Not terrific but still worthwhile as a benchmark.  This sim is remarkable in that turning off the textures doesn’t make any difference in the framerate. Its all about moving those polys, or that’s my take on the textures not making any difference.

 

Celeron 458

P-III 682e

Flanker Benchmark 1024x768 Max Details except smokestacks

10 frames a second

17 frames a second

So here is all the proof that’s needed for those of you who own 450’s and are thinking that a video card to replace your Voodoo 2 is what you need to improve your frame rates dramatically in this or in fact any Flight sim. In case its not obvious, get a new processor first…<VBG> 

Results from Janes F-15 Benchmark Mission
This is a terrific flight sim that honestly ran extremely well on my old Celery 458 processor. I just include it here cause it’s a flight sim and I can reliably bench it to within a frame or so.

Max Detail

Celeron 458

P-III 682e

Janes F-15 Benchmark 640x480

20 frames a second

30 frames a second

Results from Janes new F/A-18 Benchmark Mission
This is the newest flight simulator out and as is tradition it pushes the best computers harder than anything else. With a 3D Cockpit that absolutely sucks the life out of a 458 Celery I was really looking forward to pitting my newest beast up against the sim that’s humbling a lot of puters.

Max Detail

Celeron 458

P-III 682e

Janes F/A-18 1024x768

7-fps

11-fps

Janes F/A-18 800x600

7-fps

11-fps

Janes F/A-18 640x480

7-fps

12-fps

Alrighty then someone needs to explain why the different resolutions don’t make any difference to the Frames Per Second. This mission has lots of active objects. Yet I was surprised to find that even when I reduced the active objects a significant amount not much difference was found. 

Results from Quake III Timedemo
No test would be complete without the Mack daddy of time demo benchmarks. I ran this test on high texture details with my Voodoo 2’s using the 3dfx Open GL. (yes I know its slow but this is a comparison between CPU’s not video cards).

 

 Celeron 458

 P-III 682e

 Quake 3 High Detail 640x480

  22.1

  31.8

 Quake 3 High Detail 1024x768

  18.1

24.7 

Closing thoughts and would I do it again...?

Absolutely I would do it again. Even though I havent played too many games yet since Ive been busy writing this damn article hehe, I can tell from the Games Benchmarks that this system rocks. Do I need a video card that matches the system...oh yea. But like I said earlier I have to keep that RCS down around nothing so the wifey doesnt cut the little friends off. But there area  few benchmarks where I can tell the Voodoo's are holding me back. Was it wise to upgrade the CPU first. Well considering that in Flanker I saw a nearly 70% increase in Framerates and nearly a 55% increase in Falcon 4 framerates. I don't know of a video card upgrade over a Voodoo 2 that will give that type of increase in a Flight simulator. You need the CPU first, then you get the Video Card. So any and all advice on how to slide a GeForce DDR or Voodoo 5 past the wife will be heard. I think that with a new Video Card and my 682mhz CPU that I will see Janes F/A-18 finally getting the framerates a sim like that deserves. But still a 57% increase is nothing to sneeze at by any stretch of the imagination. Its just that I believe that my Voodoo 2 cards could be holding that one back significantly as well as Quake 3. 

Back to Part 1 Dancing the Coppermine

 

 

         Flanker Ring Homepage

List the ring sites

Next Flanker site

Next 5 Flanker sites

  Add your site

This site is Much Much Cooler and Intended to be viewed by someone using Uncle Bill's Favorite Browser Internet Explorer 4.0

Webmaster <VBG> Pierre PAPA DOC Legrand send complaints to plegrand at cox dot net

Hit Counter

All trademarks used are properties of their respective owners.
All rights reserved.  All pages Copyright © 1998, 1999, 2000 by
Pierre Legrand

Copyright of all documents and scripts belonging to this site by Pierre Legrand  1998, 1999, 2000.  Most of the information contained on this site is copyrighted material.  It is illegal to copy or redistribute this information in any way without the expressed written consent of the author.  The Author does allow items from the site to be used as a reference piece for what they are working on, but this is more for your own use, and not for public consumption or print/electronic media consumption.  This site is NOT responsible for any damage that the information on this site may cause to your system.