[rescue] FS (historical?) PPro 180 or 200, with VRM

[Curtis H. Wilbar Jr.]

On Fri, 2006-04-07 at 19:45, Mike Nicewonger wrote:
> On Apr 7, 2006, at 7:38 PM, Curtis H. Wilbar Jr. wrote:
> 
> > If prior talk on this thread is correct, fastest config for quad proc
> > for that would be 4x333 overdrives....  and if the overdrives wouldn't
> > actually go 4 way... then 4x200mhz overclocked to 233 (not by FSB, by
> > multiplier jumpers), with 1M cache each....
> 
> AFAIK PPro machines were limited to 4 way SMP. However the ALR  
> machines would do 6 way as they had some sort of ASIC that hacked a  
> way around that. That is to also say the ALR would do 6 way 333 OD  
> chips.

I wouldn't mind having a 4 way or even 6 way ppro box...  but most
seem to be pretty hulking... and noisy....  I almost bought off ebay
one time an Integraph quad ppro.. it was in a oversized tower... but
money forced me to not get it...

Speaking of PPro boxen... don't know if anyone else has tried memtest
x86 on their ppro.. but if it's memory access speed tests are to be
trusted... the ppro has workse memory access times than same mhz
pentiums (even with both having EDO memory).  This doesn't make
sense to me... and I e-mailed the author of memtest x86 but never
got any answer....

With the Ppro being a server and power workstation platform of the day,
I find it hard to believe it would have significantly slower memory
access vs a pentium... I'd expect it to have much better access.

Any insights or real world memory bandwidth experiences with ppros ?

-- Curt

> 
> --
> Mike N
> 
> So if I eat sardines does that mean I have baited breath?
> _______________________________________________
> rescue list - http://www.sunhelp.org/mailman/listinfo/rescue