The main reason why to use as less vCPU's aspossible is due to the fact that symmetric multiprocessoring is used. That affects not only VMware,it is a general virtualization issue because the physical cores are shared.
In symetric multiprocessoring the guest OS has to wait until all vCPU's are available and synchronized to schedule work to all of them. During that time the occupied cores are not available for any other task. So if you grant as much vCPU's as you have physical cores, you leave no processing power toother guests and the host as long as this particular guest resides on its vCPU's. That's never the case on bare metal, as the OS always owns all cores.
Regarding Linux this is an issue caused by older Kernels which generate up to 1000 interupts per second and additional 1000 per (v)CPU. Using other Kernels or better not using the PIT timer (which generates these interupts to get the number of ticks).
Windows uses different kinds of hardware clocks and they can change by installing software (like Quicktime).
AWo
VCP / VMware vEXPERT 2009
\[:o]===\[o:]
=Would you like to have this posting as a ringtone on your cell phone?=
=Send "Posting" to 911 for only $999999,99!=
vExpert 2009/10/11
[:o]===[o:] [: ]o=o[ :]
= Save forests! rent firewood! =
