[Shootout-list] fannkuch (timer resolution; HZ=1000?)

Bengt Kleberg bengt.kleberg@ericsson.com
Mon, 23 May 2005 11:39:36 +0200


On 2005-05-21 11:31, Jon Harrop wrote:
> On Saturday 21 May 2005 07:55, Pascal Obry wrote:
> 
>>One car argue that the JVM and the CLR are part of the design. It's not the
>>fault of compiled programs if loading the CPU cache is fast!
> 
> 
> True, but that only justifies having one benchmark concentrate on measuring 
> the startup time, not having half of the benchmarks measuring the startup 
> time.

if all bechmarks contained one measurement of the startup (for that 
benchmark) and the posssibility to subtract that value from the other 
measurments, then we would have a good comproimise. imho.

i am stilll thinking about the value for ''n'' to startup. is it 0 or 1?


bengt