COMBATSIM.COM: The Ultimate Combat Simulation and Strategy Gamers' Resource.
 

 
Voodoo3 F4 Benchmark Comparison
by Bubba "Masterfung" Wolford and Pierre "Papadoc" Legrand
 

The Problem: The Solution

If lack of standardization or visibility as a group creates these sorts of problems, then part of the solution is increased visibility and organization. Sites like COMBATSIM.COM™ are not only good sources of information, but have potential for greatly impacting the hobby we all love.

One step toward better organization and increased visibility as a group would be the creation of a standard benchmarking tool to rival the Time Demos that are in common use. For those who don't know, a beginning on that dedicated benchmark has already been made.

Pierre "PAPA DOC" Legrand has created a Falcon 4.0 benchmark that is quite good and similar to the Quake 2 "Crusher" benchmark often seen on other pages. PAPA DOC's benchmark demonstrates a scenario that offers a high degree of challenge for the video board, CPU and memory. Running this benchmark produces a good estimation of what kind of FPS you could expect to see from your own machine running Falcon 4.0 in campaign under various video configurations.

After getting my hands on a V3 3000 recently and for the first time putting it through it's paces in games WE need to benchmark, I was pleased to use the PD TimeDemo F4 benchmark. I recommend this benchmark to each website testing computer hardware and encourage them to use it as a standard for testing flight simulations.

The Numbers

Since PAPA DOC has worked so hard to get his benchmark out there, and so many of you responded by sending in your FPS numbers, we have a good diverse selection of CPU and video types to allow us to do some accurate benchmarking for different video cards (3dfx Banshee, TNT, V2 SLI, etc...) to compare to the new Falcon 4.0 video card champion, Voodoo3 3000. Click on the image below to see the benchmark comparison.

V3 Benchmarks
Click to go to Benchmarks.

Click to continue

 

Conclusion: Pierre "PAPA DOC" Legrand

First I would like to thank Bubba "MasterFung" Wolford for using my Time demo mission to illustrate the capabilities of the new 3dfx Voodoo 3-3000 Card. I guess it goes without saying that I consider the value of testing our systems with our primary application software, rather than someone else's applications, of the greatest value.

It's a shame that more developers do not follow the lead of both Quake and now Falcon in giving us the tools to tune our computer. I hope that Bubba's effort sends a message to other developers that having a log command is good for the consumer and end user.

My time demo mission is easy to run and more importantly offers the user a benchmark that will return an Average Frame rate Number within .5 tenths when repeated with identical settings. This type of repeatable test is of utmost importance to we the players since it allows us to tweak settings within either the sim or on the host machine computer and immediately determine performance changes.

Of even more importance to flight sim enthusiasts, the test utilizes the most graphically demanding simulation available today. The demands specific to an application like Falcon4 realistically represent our needs, unlike something like Quake or Forsaken.

Now for a breakdown of the frame rate test from the PD benchmark. It is apparent that all the Voodoo cards seem to do best when stressed with the combination of high resolution and high detail. Notice that the TNT products only close on the V2/V3 cards when the detail levels go down.

Oddly enough, it is also apparent that there may be some driver issues given the Voodoo 3's extremely high score using Direct X instead of Glide in one test using low resolution and low detail. This was not duplicated in any other tests where Glide always displayed higher results. This same result has been reported in Voodoo 2's when using the combo of low resolution and low detail. So it is not totally unexpected, just a greater disparity using the V3 versus V2.

I am currently revising this benchmark in cooperation with COMBATSIM.COM™ you can look for an even better benchmark soon. If all goes well we will release a cooperative effort that will become a standard for comparison of video hardware for sim fans everywhere.

Go to Page Three

 

Copyright © 1997 - 2000 COMBATSIM.COM, INC. All Rights Reserved.
 Last Updated April 12th, 1999

© 2014 COMBATSIM.COM - All Rights Reserved