OS2 World.Com Forum

Subject  :  Panorama
Author  :  melf mikaelelf@os2ug.se
Date  :  02 May, 2007 on 09:21
Just a report. I've just installed the Panorama VESA driver on my LG F122PTV, which has a dual core processor. Graphic card is Intel 82945GM. SNAP supports this card in accelarated mode but I've not been able to use widescreen resolutions. With Panorama widescreen works and the speed, as it seems for me, is just slighly slower than SNAP accelerated, performance is very good.

Subject  :  Re:Panorama
Author  :  rudi
Date  :  02 May, 2007 on 18:25
Could you give some compareable numbers (i.e. Sysbench Graphics- and DIVE test results) for both drivers ?

Subject  :  Re:Panorama
Author  :  Christian Hennecke
Date  :  02 May, 2007 on 20:21
I have an ATi X800GTO. Playing a video with WarpVision and Panorama (shadow buffer enabled) creates double the CPU load as with SNAP.

Subject  :  Re:Panorama
Author  :  e-co
Date  :  02 May, 2007 on 21:56
Hello Christian Hennecke

Please post report about your testings. PCI.EXE > pci.log, etc.


Subject  :  Re:Panorama
Author  :  melf mikaelelf@os2ug.se
Date  :  03 May, 2007 on 09:29

rudi (02 May, 2007 18:25):
Could you give some compareable numbers (i.e. Sysbench Graphics- and DIVE test results) for both drivers ?

Hard facts shows that its more slow according to SNAP accelerated than slightly, as was my feeling. I can't interpret those numbers but here they are from Sysbench 0.95:

Panorama:
PM-graphics marks: 60.471
DIVE-marks: 2073.574

SNAP:
PM-graphics marks: 269.409
DIVE-marks: 2037.804


I have the same load on Warp Vision in both drivers.

I was told at forum.ecomstation.ru that far more important scores for today's applications are:
- BitBlt S->S copy
- BitBlt M->S copy

so here are these results:

Panorama:

BitBit S->S copy 207.728
BitBit M->S copy 265.971

SNAP:

BitBit S->S copy 472.180
BitBit M->S copy 376.057

/Mikael


Subject  :  Re:Panorama
Author  :  Christian Hennecke
Date  :  03 May, 2007 on 15:14

e-co (02 May, 2007 21:56):
Hello Christian Hennecke

Please post report about your testings. PCI.EXE > pci.log, etc.


See the attachment for the pci.log.


Subject  :  Re:Panorama
Author  :  e-co
Date  :  03 May, 2007 on 16:05
empty file

we ask users follow the procedure described in tester.doc
else users send logs by fax, via anti-spam filters, to wrong emails. we receive empty files, cutted, decoded files.


Subject  :  Re:Panorama
Author  :  Glenn
Date  :  05 May, 2007 on 07:08
BTW with time passing and no additional development, SNAP is losing value. If Scitechsoft wants to sell it, they'd better give it a price decreasing with time or keep improving it by adding support for new adapters...

Subject  :  Re:Panorama
Author  :  rudi
Date  :  06 May, 2007 on 09:00

melf (03 May, 2007 14:21):

Panorama:
PM-graphics marks: 60.471
DIVE-marks: 2073.574

SNAP:
PM-graphics marks: 269.409
DIVE-marks: 2037.804

I have the same load on Warp Vision in both drivers.


Not really unexpected. Even though you didn't mention most interresting "video bus bandwidth", the DIVE performance of your system is excellent. However, for movie watching overlay support would still be desirable to have overlay support. I guess Panorama will not offer this feature...



melf (03 May, 2007 14:21):

Panorama:

BitBit S->S copy 207.728
BitBit M->S copy 265.971

SNAP:

BitBit S->S copy 472.180
BitBit M->S copy 376.057


Event though we see the difference, the absolute values suggest that it's quite useable. I guess this is mostly due to the very good overall performance of your system. I wondering, if you get any visible performance degration when scrolling text in a command window...


So just like any other unaccelerated solution, Panorama is better than nothing, but not great eather. In your specific situation, you will most likely prefer it over SNAP, due to the aspect ratio / resolution problems.


BTW, here are benchmarks of a 5 years old system (Athlon XP 2100+) running SNAP on an nForce integrated graphics (GeForce2 compatible):

BitBlt S->S copy : 420.256
BitBlt M->S copy : 236.919
Filled Rectangle : 1726.477
Pattern Fill : 1722.214
Vertical Lines : 44.786
Horizontal Lines : 866.480
Diagonal Lines : 23.601
Text Render : 292.408
------------------------------------
Total : 307.184

Direct Interface to video extensions -
Video bus bandwidth : 493.884
DIVE fun : 1749.602
M->S, DD, 1.00:1 : 1684.800
------------------------------------
Total : 634.315


Subject  :  Re:Panorama
Author  :  rudi
Date  :  06 May, 2007 on 09:05

Glenn (05 May, 2007 07:10):
or keep improving it by adding support for new adapters...

I don't think there is - besides Kendall Benett himself - any person left at Scitech to work on the code. And Kendall made clear, that he has/wants to do other things...


Powered by UltraBoard 2000 <www.ub2k.com>