The pinning ends up being essentially free. There is no impact on performance. You are correct. Nio becomes very usefull when the memory associated with a primitive is controlled by native code, such as in some current nvidia extensions. Today we use vertex arrays with no extensions (or display lists), so data movement is completely up to the driver. Whether or not it is serial depends somewhat on the graphics card. Efficient usage of some of the new data movement extensions is also very application dependent. We are looking at ways to incorporate it into the API though.
Doug. Artur Biesiadowski wrote: > Doug Twilleager wrote: > >> If all of your data is created in java land, then nio doesn't get you >> any faster. If you read the current specifications about JNI, it talks >> about the potential of data being copied when it is passed from java >> through JNI. There are API's in JNI that prevent such a copy. > > > But it pins arrays (gc-wise), doesn't it ? Not that gc has a lot to do > in normal java3d app, but... > > I imagine that nio would also be a big win if asynchronous loading of > data could be utilized (NV_vertex_range ?). Does it mean that currently > all data passed from java3d to graphic card is fed in serial manner ? > > Artur > > =========================================================================== > > To unsubscribe, send email to [EMAIL PROTECTED] and include in the > body > of the message "signoff JAVA3D-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 JAVA3D-INTEREST". For general help, send email to [EMAIL PROTECTED] and include in the body of the message "help".
