[EMAIL PROTECTED] wrote:
processing in software could benefit from porting
their operations to
the GPU -- regardless of whether those pixels are
going straight to the
display or coming back out of VRAM to be written to a
file.
Yes I know gpgpu and yes I've already on such a project, I also have some posts
in their forums.
Now write some software which does what you describe here, implement it using
the D3D pipeline including readbacks to VRam and once with the headless
software-only Toolkit - now load your service with 20-50 concurrent users
(otherwise the faster pipeline would not make any sence) - and compare your
results.
No matter wether you use a 8800GTX or a Readeon-3870, the results will be
disappointing compared with the software-only pipeline - if you don't believe
me simply try it out.
I agree - for servers it's much easier to scale using CPUs
than GPUs - at least, for now.
Thanks,
Dmitri
lg Clemens
[Message sent by forum member 'linuxhippy' (linuxhippy)]
http://forums.java.net/jive/thread.jspa?messageID=246354
===========================================================================
To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
of the message "signoff JAVA2D-INTEREST". For general help, send email to
[EMAIL PROTECTED] and include in the body of the message "help".
===========================================================================
To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
of the message "signoff JAVA2D-INTEREST". For general help, send email to
[EMAIL PROTECTED] and include in the body of the message "help".
