Keep 3DLuVr online!
3DLuVr Logo
 From the Real World
 Digital Painting Series
 Featuring of...
 On the Bookshelf
 3ds max
 Softimage XSI
 Rhinoceros 3D
 Video Tutorials
FunZone menu
 I always wanted to be
 Talk to an employer
 Why Ask "Why"
TechZone menu
 Hardware Reviews
 Software Reviews
 Q&A, Tips & Tricks
UserZone menu
 The Artist Sites
 15 Min of Fame
 Request an Account
 Current Assignment
 Sponsors & Prizes
 Make a Submission
 Voting Booth
 Competition Rules
About menu
 Mission Statement
 Poll Archive
 How to IRC
Log in to be able to post comments to the news items, forum posts, and other facilities.
Not registered? Register!     Lost Password?
 Your New Year`s Resolution is...
Gain employment
Stop smoking/drinking/etc
Get back in shape
Find the meaning of life
Conquer the World
Absolutely nothing

    Poll Results
Want to leave us a comment about the site or in general? Click here to access the form.
TechZone Heading
The Celeron Myth: Real World 3DSMAX Performance
Added on: Wed Jun 16 1999
Page: 1 2 4 

Average Viewport FPS

Dolphins.max (67,000 poly animation from above)
Bench1.max (3 objects totaling 20k polygons lit by two spotlights)
Bench6.max (One 20k polygon object never clipped by the viewports)

By setting ShowFPS=1 in the 3dsmax.ini file under [Performance], you can get a readout of the frames per second in the currently selected viewport window.
Using this knowledge, I used three main files to show the performance of 3DSMAX R2 Software HEIDI drivers over a wide spectrum.

For each benchmark, adaptive degradation was turned off, the perspective view was made full screen, and full textures/smoothing was enabled. Real mode animation preview was turned off, and 3DSMAX's preview animation button was used to show the ability to fill raw polygons. Screen captures were taken at frame 20 to show the most consistent Frames Per Second for each benchmark. This process was repeated multiple times, and an average value for each benchmark was found. For this graph a higher bar indicates
a higher average frames per second.

Viewport FPS benchmark

Viewport Animation

Dolphins.max, Bench1.max, Bench6.max

This test was performed similar to the Average Viewport FPS test, with the exception that it was a timed benchmark. From the beginning of the animation to the end was timed with a stopwatch for at LEAST three repetitions. The average time was then taken.
The two viewport tests help to show the possible differences between the Celeron and Pentium II systems in general viewport modeling. A lower bar in this graph indicates a faster animation result.

Viewport Animation


Is a Dual Celeron System a Viable Alternative to a Dual Pentium II System?

The answer is...yes. The Dual Celeron system performs just as well if not better then a similarly equipped Pentium II system. Contrary to popular belief, adding a second Celeron increases your performance in 3DSMAX rendering around 80-90% and your viewport modeling around 35-40% JUST like a Pentium II system. It would appear that even under the harshes of circumstances, the 128k on chip L2 cache beats the half clock speed 512k L2 cache of the Pentium II. Across all fourteen benchmarks the Celeron shows an average of a 1.6% increase in speed over a Pentium II. While that doesn't seem to be a noticable amount, you must realize that two Celeron 300a's (PPGA) cost about 100 dollars,
while two Pentium II 450's cost upwards of 400 dollars.

1997-2017 3DLuVrTM (Three Dee Lover)
Best viewed in 1024x768 or higher,
using any modern CSS compliant browser.