Takedasun wrote on 2023-12-11, 21:41:
Oh! Thanks a lot! That's great!
The driver is newer(4338) than mine(4324). The resolution issues keeps the same unfortunately. However the 3D performance increased obviously. I've done a 3D testing through GLQuake timedemo1. Here is the results (I set the MPact 3D Performance to Auto Select through out the whole test):
1st test (with 4338 driver):
800x600-16bpp: 8.4fps
640x480-24bpp: 12.1fps (Darkened color)
640x480-16bpp: 12.8fps (Darkened color)
320x240-16bpp: 37.2fps
* 800x600-24bpp and 320x240-24bpp have display issues, the former leaves some part of image out of the top/bottom boundary and the later results in the image limited at the left corner of the screen.
2nd test (with 4324 driver):
800x600-16bpp:4.0fps
640x480-16bpp: 5.2fps (Darkened color)
640x480-32bpp: 8.1fps (Darkened color)
320x240-16bpp: 32.3fps
After this, I thought the results may be affected by the heat accumulation on the card. So I decided to do a second test. And the results are:
3rd test (with 4338 driver):
800x600-16bpp: 8.3fps
640x480-24bpp: 10.8fps (Darkened color)
640x480-16bpp: 12.5fps (Darkened color)
320x240-16bpp: 32.9fps
The result proved my thought. The performance does reduced, but it still a lot better than with driver 4324. So I decide to make an extra test just to see the result with more heat.
4rd test (with 4338 driver):
800x600-16bpp: 7.4fps
After so much comparison, I think the new driver indeed increased the 3D performance of the card.
This guy got a better performance than mine. He used the 4333 version which I don't have. But I think he might set the 3D performance slider to the max, hence the result.