From: Allan Schaffer (allan++at++sgi.com)
Date: 06/01/2000 18:25:12
On Jun 1, 4:20pm, Luc Leblanc wrote:
> <p>Let's say we have an Onyx 2 IR2 with 1 pipe (<i>2 RMs)</i> driving driving
> two channels at 1280x1024 (<i>medium pixel depth: 512 bits per pixel</i>)
> and another Onyx 2 IR2 with 1 pipe (<i>4 RMs) </i>driving two channels
> at 1280x1024 (<i>large pixel depth: 1024 bits/pixel</i>). Would there be
> a difference in performance (framerate) when the same software is run on
> both machines ?
Hardware-wise the 4RM system will have twice the fill rate as the 2RM
system -- quite a difference. You'll only notice this if you're fill
limited of course, but most apps are.
There are no specific performance differences between otherwise
identical systems offering 'medium' and 'large' pixel depth per se,
but a system reporting large pixel depth has more OpenGL visuals
available than one with medium pixel depth, which means that
Performer might end up using a better quality OpenGL visual on the
'large' system than on the 'medium' system. For example, perhaps you
can get 8 multisamples with large, but only 4 with medium.
If this is the case, the lower-quality choice will generally be
faster. To quantify how much faster, you'd need to find out exactly
which visuals are used (or set them yourself).
Allan
ps. please send messages as plain text, not HTML
-- Allan Schaffer allan++at++sgi.com Silicon Graphics http://reality.sgi.com/allan
This archive was generated by hypermail 2b29 : Thu Jun 01 2000 - 18:25:20 PDT