Maybe you CPU folk will be able to help me with this Hydravision transparent windows/CPU usge question ;)

[dug777]

At least half as high a CPU usage jump dragging the taskmanager window about my dual monitor desktop when transparency is enabled as i do when transparent dragging windows is disabled....

is it offloading the work more onto my 9800 pro?

Cheers folks,

Dug.

( sorry about the massive cross posting but no-one in Video seems interested )

 

[ProviaFan]

Are you saying that CPU usage is higher when using transparent dragging, or lower? Taken literally, your sentence seems to indicate that CPU usage is lower while transparent windows are being moved, but that has never been my experience in practice. Theoretically, the video card has the ability to accelerate 2D blending operations, but I've never seen this utilized, at least on ATI cards.

 

[dug777]

Originally posted by: ProviaFan
Are you saying that CPU usage is higher when using transparent dragging, or lower? Taken literally, your sentence seems to indicate that CPU usage is lower while transparent windows are being moved, but that has never been my experience in practice. Theoretically, the video card has the ability to accelerate 2D blending operations, but I've never seen this utilized, at least on ATI cards.

thats exactly what i'm saying :thumbsup:

At least half as high as when i disable transparent window dragging in Hydravision and drag a window around my duallie desktop(my test window being task manager). I've tried it dozens of times to make sure, and it's definitely the case...

 

[MDme]

half as high vs NON-TRANSPARENT windows???

 

[dug777]

Originally posted by: MDme
half as high vs NON-TRANSPARENT windows???

yeah.

for example, if i drag the taskmanager window (or a msn chat window) normally around my dual monitor desktop my CPU usage will hit as high as 86% and stay high until i release the window. But with transparent window dragging enabled in Hydravision it doesnt usually peak over 25% initially, then falls back to about 7-8% as i continue dragging, occasionally spiking up to 25% again as i continue...

and it is repeatable time after time, no matter which window i choose to drag.

 

[MDme]

My guess is, when you use transparent window dragging the system uses accelerated GPU routines, where as "simple" non-transparent window dragging may be using more "traditional" routines.

 

[dug777]

Originally posted by: MDme
My guess is, when you use transparent window dragging the system uses accelerated GPU routines, where as "simple" non-transparent window dragging may be using more "traditional" routines.

Cheers, that sounds like it

 

[MDme]

glad to be of help

 

[dug777]

anyone else got any ideas?

can anyone else try this out?

 

[dug777]

bump for the morning crowd

 

[mircea]

Well I just tried this on a single monitor and I didn't get the same tipe of action. I get a top usage of 7% with transparency and 4% without it. I also have the performange settings for windows set on best performance.
I'll try dual monitor when I'll bring down my brothers and let you know what I see.

 

[dug777]

Originally posted by: mircea
Well I just tried this on a single monitor and I didn't get the same tipe of action. I get a top usage of 7% with transparency and 4% without it. I also have the performange settings for windows set on best performance.
I'll try dual monitor when I'll bring down my brothers and let you know what I see.

are you enabling transparency through Hydravision? If so i'm not surprised you see no benefit given it's written for ATI cards AFAIK... (are u using that 6600GT on your rig page?)

It would be good if you could check with the duallies tho...

as i said my results are repeatable time after time so i'm pretty sure it must be offloading something onto the GPU. (9800 pro).