Orbit in 2D wireframe takes too long to initiate and performance in 3D wireframe is poor.
Initiating the orbit command in 2D wireframe takes 30 seconds and then if most of the layers are on it then regenerates the model and takes another 40 seconds to initiate, a total of a minute. To end the orbit in 2D wireframe it takes 20 seconds and again if most of the layers are on it takes another 10 seconds to regenerate, a half minute; it takes nearly two minutes to initiate and end each orbit. In 3D wireframe and conceptual the orbit command is nearly instant but the orbit, pan, and zoom commands are much slower than in 2D wireframe (enough that I need to use 2D).
For comparison my 2006 HP dv2120 laptop with a Turion x2 2.3Ghz, 8GB RAM, Muskin SSD, and an integrated GeForce GO 6150, initiated the 2D wireframe nearly instantly but had trouble initiating the 3D wireframe. For testing I have tried a Quadro FX 5600, several different drivers, and changed various settings but nothing helps the 2D wireframe orbit.
Your experience, knowledge, and intuition is appreciated.
Open-hardware workstation
AMD Deneb C3 x4 3.6 GHz
ASUS M4A87TD (non-EVO) AMD 870|SB850
16 GB DDR3 RAM 4x4 PC12800/1600MHz
OCZ REVO Drive 1RVDX0240
Liquid cooled
Server 2008R2
C3D 2012 F.51.0.0
FirePro 3D V7800 2GB RAM Driver v8.773.1.1 Server2008
Solved! Go to Solution.
Solved by granite07. Go to Solution.
increase memory cache size reduced 3D orbit command initiation time from 18-seconds to 5-seconds
http://www.ghacks.net/2010/07/08/increase-the-filesystem-memory-cache-size-in-windows-7/
Are you absolutely sure that's what made a difference?
According to this, that key is in the registry purely for backward-compatiblity, and is not referenced anywhere by any part of Win 7:
http://www.tweakhound.com/blog/?p=1164
yes
documentation of the test conditions and procedures is here
Wow... That's an awful lot to sort through.
But it looks like you failed to do the simplest test: change that setting back to 1 and see if it has any effect. You should see the problem come back, if this is really that setting that did it. Then change it back to 2, and verify the problem goes away again. This doesn't necessarily confirm that the problem is affected by that setting, but it will give you a fair bit more confidence.
It just seems weird that something that is in no way referenced by the OS should have any effect...
As a side note, I also saw your question about the difference between "C3D Imperial" and "C3D Metric". The answer is that there aren't two "flavors" of C3D; the only difference is the user Profile used during startup. That's it. There's no other difference. And for the most part, the only difference between the user Profiles is which template is set as the QNew template. I actually can't think of anything else that's different.
I am convinced that no further testing in necessary to validate and defend; also I do not want to take any chances that I will unfix my machine from the present 'good' performance.
Lets see if anyone else can duplicate the results as a better validation than if my specific machine toggles between 'barely works' and 'works really good'' with the flip of a digit. There are lots of posts for those with slow regen times and orbit initiation - post your results here.