On Sat, 12 Nov 2005 03:53:46 +0000, Roy Schestowitz wrote:
> __/ [Linønut] on Saturday 12 November 2005 01:25 \__
>> After takin' a swig o' grog, Rich Gibbs belched out this bit o' wisdom:
>>> Gordon Burgess-Parker said the following, on 11/11/05 14:48:
>>> Here's a link to the original Microsoft paper (.pdf):
>>> I haven't had a chance to read it yet, but some of the performance
>>> comparisons are, um, interesting.
>> I like this (from the blog):
>> So why is this interesting? Because their test methods reflect
>> Windows internals, not Unix kernel design. There are better, faster,
>> ways of doing these things in Unix, but these guys - among the best
>> and brightest programmers working at Microsoft- either didn't know or
>> didn't care.
>> And if they're the best and brightest, what do you think happens when
>> the average Microsoft programming whiz gets asked to program for
> Microsoft finally *give* the fact, but hide them under a thick blanket (PDF)
> that would better not leak to the world:
> What's noteworthy about it is that Microsoft compared Singularity to FreeBSD
> and Linux as well as Windows/XP - and almost every result shows Windows
> losing to the two Unix variants. For example, they show the number of CPU
> cycles needed to "create and start a process" as 1,032,000 for FreeBSD,
> 719,000 for Linux, and 5,376,000 for Windows/XP.
There are no surprises here. What these tests quantify is what has been
known for a long time. Process creation on Windows is expensive, while
thread creation and manipulation is cheap. Windows was designed with that
philosophy in mind.
Unix, on the other hand, for a long time didn't have kernel threads, and
therefore a lot of work was done to optimize process creation. Some Unix
now have kernel threads, but Windows still outperforms them in thread by
Also, "almost every result" appears to mean 50%, since 3 of the results are
better than Unix/Linux, while the other three are not. Kind of a