> I'm aware of the CPU difference and I would have thought that 2 modules
> sharing 80Mhz would statistically perform better than 4 modules sharing
> 100Mhz. I'm also aware of the memory capacity difference and again that
> statistically, 2 modules sharing 64M wouldn't be much worse than 4 modules
> sharing 128M (if your application even requires maxing out the memory).
> Can you be more specific on the bus bandwidth difference that you mention?
> And, do you mean memory capacity or is there actually a memory bandwidth
> value that is important? Any performance stats would be appreciated.
Except that the amount of memory required isn't huge per module, but is
huge when running BGP. And as some have pointed out, the 64mb boxes
will stop being able to take full BGP tables sooner than 128mb boxes.
This archive was generated by hypermail 2b29 : Sun Aug 04 2002 - 04:13:16 EDT