"Where there is love there is life."
- Mahatma Gandhi
More pages: 1 2 3 4
New dynamic branching demo
Tuesday, April 25, 2006 | Permalink

There's a new demo showing off the benefits of dynamic branching. The demo should run on Radeon 9500 and up and GeForce FX and up, but performance improvements by enabling dynamic branching is only expected to happen on Radeon X1300 and up and possibly also GeForce 6 series and up (haven't tested) since earlier chips do not support this feature.

Name

Comment

Enter the code below



Dan
Monday, May 8, 2006

There are unfixed things in ATI drivers by more than one year. There is also the utter ignorance of ATI to the fact thats FBOs are futile for anyone doing stencil shadows whithout EXT_packed_depth_stencil.

There are also some more indeep flaws in ATI drivers, which aint obvious.

I think the job the ATI driver team is lousy and ATI lacks consideration for his customers

Humus
Tuesday, May 9, 2006

To begin with, this is my personal site. If you have comments about the demos feel free to share, but this is no place for general whining about the drivers. Also, issues are naturally resolved in order of importance. But what you find important may not be top priority to other developers.

Dan
Tuesday, May 9, 2006

This wanst a whining. I just found another ATI person through which I can tell ATI driver team they suck big time. This is not a whining , nor a suposition, is simply a conclusion. Also some of your engineeres dint even realize that EXT_packed_depth_stencil is required to work with FBOs and stencil shadows. Lousy, aint it ? I wonder how they got hired in the first place

Yes is your personal site, but you have many times indirect borderline comments about Nvidia hardware, so I decided to make a comment as well about ATI. Im sorry if it offended you personaly, it wanst my intention.

Quasar
Tuesday, May 9, 2006

Several users over at the german 3dcenter-forum report that they get increasingly higher avg. Fps whilst reducing their CPU clockspeeds.
If you want to have a look at the numbers: http://www.forum-3dcenter.de/vbulletin/showthread.php?p=4287553#post4287553 (if you understand german of course you can do more than look at the numbers.)

Any idea how that's possible?

AnarchX
Tuesday, May 9, 2006

direct link to my results of the cpuclock-anomalie, which only concern nvidia cards:

http://www.forum-3dcenter.org/vbulletin/showpost.php?p=4287762&postcount=42

fellix
Monday, May 15, 2006

I got a nice little boost in average FPS with Branching enabled after installing the recently leaked 87.25 drivers on my 6800GS.
Can anyone confurm on GF7 class card?

Humus
Monday, May 15, 2006

AnarchX, the only reason I can think of would be that the timers on your system screw up in some way. I don't think it's actually running faster. Could you compare with for instance Fraps and see if there's any difference in the reported framerate?

AnarchX
Tuesday, May 16, 2006

3000MHz
2006-05-16 13:12:03 - DynamicBranching3
Frames: 4176 - Time: 20000ms - Avg: 208.800 - Min: 182 - Max: 255

1500MHz
2006-05-16 13:12:59 - DynamicBranching3
Frames: 7858 - Time: 20000ms - Avg: 392.900 - Min: 346 - Max: 457

It look like that the movements in the demo get slower if I decrease my cpuspeed.
Maybe a timer in the OpenGL-Part of the Nvidia-driver is synced with the win-start-clock?

More pages: 1 2 3 4