Video RAM Caching

 






Expendable (640x480x16)

  Enabled Disabled
Minimum 29.0 29.3
Maximum 91.7 90.0
Average 48.8 49.1

When the caching of video RAM was enabled, the framerates recorded had a slightly larger variance of 62.7 fps - that's for a high of 91.7 fps and a low of 29.0 fps. When the caching was disabled, the framerates were a little more stable with a variance of 60.7 fps. With the minimum a little higher at 29.3 fps, the disabled setting managed to score a higher average framerate of 49.1 fps - 0.3 fps or 0.6% higher.

 

Expendable (800x600x16)

  Enabled Disabled
Minimum 29.0 29.3
Maximum 80.3 82.3
Average 48.5 48.4

At the resolution of 800x600, the disabled setting now had a wider variance. The framerates it recorded had a minimum of 29.3 fps and a maximum of 82.3 fps. Thus, there was a 53 fps variance. The enabled setting, on the other hand, was a little more stable with a variance of 51.3 fps.

Although the disabled setting had both the higher minimum and maximum framerates, it lost in the average framerate by 0.1 fps.

 

Expendable (1024x768x16)

  Enabled Disabled
Minimum 22.3 26.3
Maximum 70.7 70.3
Average 45.6 45.6

When the resolution increased to 1024x768, caching the video RAM surprisingly decreased minimum framerate by 18%! But caching also enabled the enabled setting to beat the disabled setting by 0.4 fps in the maximum framerate. The variance was higher for the enabled setting at 48.4 fps while the disabled setting's variance was only 44 fps.

On the average though, both settings scored rather similarly though the disabled setting was slightly ahead. As we can see, disabling the feature allowed for a less variable (more stable) framerate that also has a higher minimum framerate, allowing for less jerkiness. However, as the resolution increased, the performance advantage of disabling this feature decreased.

 

Turok : Dinosaur Hunter

  Enabled Disabled
FPS 132.83 133.50

In this single-texturing game, enabling the caching of video RAM decreased the average framerate by 0.7 fps or about 0.5%. In all three test runs, the disabled setting scored both the highest maximum and minimum scores. So, I would say that the disabled setting is the clear winner here.

 

 

 
 

 

 
     
   

 

 
   

 
     
 

                   

 
   

 

 
 
Last Updated 15-09-2000

All trademarks used are properties of their respective owners.
Copyright © 1998-2000 Adrian Wong. All rights reserved.

 
Visit the new Tech ARP @ http://www.techarp.com/ !