This paper shows that it is possible Java to have better general performance than C++, and that it is likely that this will actually occur. Java performance can exceed that of C++ because dynamic compilation gives the Java compiler access to runtime information not available to a C++ compiler.

Java bytecode + dynamic compilation *WAS* the idea JIT was based on. The article is apparently old enough not to take into account concepts that came in later with HotSpot, etc. Seriously, those we all dreams. Big dreams that looked great on paper, but never came through, and likely never will considering the fate of Sun. And before I get a you-dont-get-it, I must say I was actually a Java guy in my younger days (even wrote my Master thesis in Java). As for interpreted language speed considerations, I’m very much aware of those and the conclusion I came to in the end is that no JVM magic wand can help as much as the additional brainpower I can throw at the problem, and that’s why I prefer Python nowadays, with C(++) for the heavy artillery (the ability to use Qt from both makes this very sweet). Long story short, you don’t have to convince me, been there, done that, said goodbye to Java