Rick Kelly writes: 

> Perry E. Metzger said: 
> 
>>Yes there is. The locking protocols on the two processors are not
>>guaranteed to have compatible bugs if the processors aren't the same
>>stepping. I'm pretty surprised that this is working for him --
>>everything I know for x86 says if you do SMP, you need identical
>>processor steppings.
> 
> Sequent got away with it on their 386 boxes, but the 486 machines did
> use uniform processor speeds, as I remember. 
> 

AFAIK...the Sequents were closely coupled SMP on a given board (and always 
had the same CPU speeds on a given board), but between boards it was a 
loosely coupled MP affair.  Thus it was easier to have boards of different 
speeds, as stepping issues, etc., didn't come into play. 

Of course, I could be wrong... 

Anyone with really good memory remember if the Balances had different CPU 
speeds? 

Ken