Peter's slides (was Kaffe's JIT mechanism description)

[Enno Rehling]

>Also, forgive me for being a bit sarcastic here, the custom edition seems
>to have made another breakthrough in JIT technology.  To perform 3,000,000
>function calls in 205 milliseconds, as claimed in
>http://www.transvirtual.com/users/peter/one4all/c-performance.html
>that is, one function call in 68 nanoseconds on a Pentium/233, is orders
>of magnitude faster than what I measured.   ;-)


I was lucky to see these slides at the JavaDays'98 in Frankfurt today, and
he explains that. The test was done with a customized VM, the array bounds
checking and a few other time consuming mechanisms had been stripped. Plus,
the C code he compares it too wasn't optimized, thus results where the Java
is faster than C. It got quite a few sarcastic murmurs from the people
around me. I personally wouldn't have shown that slide.

BTW, Peter: A good presentation, even if many people in the audience clearly
couldn't make sense of the slides ;-) The fonts are just too small, and if
you've never seen the inside of a VM, it's hard to make sense of it all. But
then again, if you can't make sense of the slides, you're not the targeted
audience anyway. ;-)

Enno.