On 11 Jul, Pawel Krawczyk wrote:

> Is the difference in CPU power really that big or is this result of
> architecture differences that biases the test? I can also guess that
> the x86 implementation was heavily optimized, but anyway it's almost
> 4 times...
VAXen are slow, but they feel fast. PeeCees of the same vintage are
fast, but feel slow. That is mainly due to the suboptimal hardware
architecture of PeeCees. The other reason are some x86 assembler
optimizations in OpenSSL, notably for rc[45], sha1, md5, ... 
-- 



tschüß,
         Jochen

Homepage: http://www.unixag-kl.fh-kl.de/~jkunz/