G3 Upgrade Questions....

[Gabriel Paubert]

On 29 Jul 1999, Stefan Berndtsson wrote:

> 
> Geert Uytterhoeven <Geert.Uytterhoeven at cs.kuleuven.ac.be> writes:
> 
> > On 29 Jul 1999, Stefan Berndtsson wrote:
> > > "Timothy A. Seufert" <tas at mindspring.com> writes:
> > > > As a general rule the 750 scores about 2 bogomips per MHz, so this is
> > > > suspicious too.
> > > 
> > > On my Lombard/333 I get:
> > > 
> > > # cat /proc/cpuinfo
> > > processor       : 0
> > > cpu             : 750
> > > temperature     : 0 C
> > > clock           : 333MHz
> > > revision        : 130.2
> > > bogomips        : 333.41
> > > zero pages      : total 0 (0Kb) current: 0 (0Kb) hits: 0/256 (0%)
> > > machine         : PowerBook1,1
> > > motherboard     : PowerBook1,1 MacRISC Power Macintosh
> > > L2 cache        : 512K unified
> > > memory          : 64MB
> > > 
> > > so either it's reporting things wrong for me, or 750 doesn't always score 2/MHz.
> > 
> > Since the 750 is derived from the 603e, it scores 1 bogomips per MHz. The 604e
> > does 2 bogomips per MHz.
> 
> That's not true either.. a friends G3/300 (some stationary thing), reports 601.xx
> bogomips.. unless something's screwed with _that_ setup.

Can't alignment of the BogoMips loop affect this timing loop ? It does
on 603e AFAICT. 

OTOH, the 750 is only loosely based on the 603e. It's much faster on tight
loops because it has dynamic branch prediction (which does not help much
in this case) and an associated branch target cache which holds the first
2 instructions at the target of a branch (which is very important in this
case). However my 200MHz 603e report 133 BogoMIPS (2/3 of clock
frequency).  
	
	Gabriel.


[[ This message was sent via the linuxppc-dev mailing list.  Replies are ]]
[[ not  forced  back  to the list, so be sure to Cc linuxppc-dev if your ]]
[[ reply is of general interest. Please check http://lists.linuxppc.org/ ]]
[[ and http://www.linuxppc.org/ for useful information before posting.   ]]