OpenGL Convolution & Histograms question

OpenGL Convolution & Histograms question

Post by Loren Knutso » Wed, 23 Oct 1996 04:00:00



I've got a quick question concerning the convolution & histogram
extensions under the SGI implementation of OpenGL.  I'm using RE3
hardware under IRIX 6.2.  The routines don't seem much faster (if at
all) than the pure CPU counterparts that I coded up.  Does anyone know
if these two extensions actually use the graphics hardware or are they
implemented using the CPU instead.

Thanks in advance.

--
 Regards,
   Loren Knutson

 
 
 

OpenGL Convolution & Histograms question

Post by Allen Ak » Wed, 23 Oct 1996 04:00:00



| I've got a quick question concerning the convolution & histogram
| extensions under the SGI implementation of OpenGL.  I'm using RE3
| hardware under IRIX 6.2. ...

Do you mean InfiniteReality (IR)?

|                      ...  The routines don't seem much faster (if at
| all) than the pure CPU counterparts that I coded up.  ...

It wouldn't surprise me greatly to hear of a situation in which
host-based histogramming is faster than histogramming in the pipe, if
you're using very small images (64x64 or less) or monochrome images
with large histogram tables (greater than 4K entries).  The cost of
moving the image into the pipe and the histogram table out of the pipe
might exceed the cost of performing the histogram on the host.

It does surprise me to hear of a situation in which host-based
convolution is faster than convolution in the pipe.  (Unless you're
running into a data-transfer bottleneck of some kind, or you're
waiting for the convolution results to be read back to the host.)

|                                                   ... Does anyone know
| if these two extensions actually use the graphics hardware or are they
| implemented using the CPU instead.

On IR, there's essentially no fallback to software; nearly everything
is in either silicon or microcode.  Both histogramming and convolution
are implemented by microcode and hardware assist in the GEs.

Without knowing more about your code, it's hard to say what's
happening.  If you're processing small images and waiting to read back
the results, performance might be dominated by pipeline latency.  If
you're processing large images, then some mode setting might be
suspect.

Also, does your host-based convolution code use floating-point
arithmetic?  If not, the pipe may be performing the convolution at a
higher precision than you need.

Allen

 
 
 

1. OpenGL Convolution & Histograms question

I've got a quick question concerning the convolution & histogram
extensions under the SGI implementation of OpenGL.  I'm using RE3
hardware under IRIX 6.2.  The routines don't seem much faster (if at
all) than the pure CPU counterparts that I coded up.  Does anyone know
if these two extensions actually use the graphics hardware or are they
implemented using the CPU instead.

Thanks in advance.

--
 Regards,
   Loren Knutson

2. IEEE Visualization '94 Conference, October 17-21, Washington, DC

3. Simple Question about RAT 5

4. OpenGL Convolution performance on O2 (or lack therof)

5. graphics

6. Convolution in OpenGL

7. 3D-Animator wanted

8. Example of convolution in OpenGL?

9. convolution limits in OpenGL imaging

10. Convolutions in OpenGL

11. Example of convolution in OpenGL?

12. Real Time Convolution - timing question