Carlos Santiago wrote:
>Observing the data in annex, they do not agree that one VM with version
>1.3 or 1.4 has a better performance of what one 1.5? Another detail,
>hardwares is different yes, and knows that this influence, but to the
>point of the same program in "supposed" a version of VM to be until 10x
>slower? It does not sound as strange?
I assume the "10x slower" you refer to is the 136 ms for Java 6 on your
iBook. If I understand your point, then yes, that does seem strange.
There is an explanation, but I can't describe it directly.
To understand what is happening, you must first read the release notes for
Java 6 DP 4. You can read them online by logging into ADC, clicking
"Downloads", then clicking the link for "Release Notes (HTML)" under the
Java 6 DP 4 heading.
With the release notes displayed, search the text for "JIT", and read what
it says about the JIT compiler on PowerPC machines. If you already know
how Java improves its speed, then that one statement should be enough to
explain the performance difference you see on your iBook.
If you don't understand the significance of a JIT compiler to Java speed,
feel free to ask a question on this list.
If you ask a JIT question, please avoid saying anything about Java 6 DP 4,
because there isn't supposed to be any public discussion of Java 6 DP 4:
it's under NDA. If Apple wanted to, they could probably stop all
discussion on this right now, and remove everyone who has contributed to
this thread from the list. I'm hoping they don't do that as long as we
avoid talking about DP 4 in specific terms.
So if you have a question about Java's JIT compiler and how it affects
speed, please ask it in general terms and we'll do our best to answer it.
-- GG
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Java-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/java-dev/email@hidden
This email sent to email@hidden
