[android-developers] What is the advantages of dynamic decisions of Renderscript?

I read an article which was written by Tim Bray.

In the article, there is a statement,
"A secondary effect of choosing runtime thread-launch management
is that dynamic decisions can be made about where to run a script"

I know Renderscript JIT compiler in honeycomb caches target-based
binary files and reuse it if the previous cache files exist.

Then, I wonder whether or not Renderscript has dynamic decisions about
where to run a script on CPU vs GPU whenever it runs..
and, do the decisions have relations to current remaining resources?
e.g. the number of CPU cores, CPU frequency, CPU or GPU usage, etc.

If the decisions are related to current remaining resources,
I think that caches to reduce loading time are not needed.

The reason is because each binary may partially run on CPU or GPU,
which means target-based binary files to run are dynamic.

In conclusion,
I anticipate as the followings.

1. the dynamic decisions exists only for optimally target-based back-
end compilations.
2. the target-optimized binaries are cached on the devices at the
first running in a device.
3. after the first running in a device, the dynamic decisions are not
occurred and the cached binaries are reused.
4. Renderscript do not use scheduling based on real-time available
resources.

Those are just my suppositions.

If you have another idea or know some routines for Renderscript,
let me know it.

Thanks.

--
You received this message because you are subscribed to the Google
Groups "Android Developers" group.
To post to this group, send email to android-developers@googlegroups.com
To unsubscribe from this group, send email to
android-developers+unsubscribe@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/android-developers?hl=en