On Apr 3, 2:45 am, blindfold <[email protected]> wrote: > Thank you, Koji, these are interesting figures. Apparently the Java > interpreter for Dalvik bytecodes adds two to three orders of magnitude > computational overhead. Do you think that this is avoidable through a > further engineered redesign of your Java Dalvik VM?
You're racing a JIT-compiled Java implementation against a hand- written assembly implementation of the very same thing. The asm version is going to win. A more interesting race would be to see if the JIT-compiled library implementations of the underlying VM provide a speed boost. A benchmark that did a lot of string manipulation might end up being closer because it spends more of its time in library code that the interpreter doesn't touch. --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Android Discuss" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/android-discuss?hl=en -~----------~----~----~----~------~----~------~--~---
