> > [The 11/730] was the slowest VAX ever shipped, with the only possible
> > exception being the MicroVAX-I.

The uVax1 was indeed slower than a Vax 725/730 but that was more due to
I/O and memory devices than to the CPU.  If the uVax1 had had a cache,
it would have been as fast at most things as as a 725/730 -- basically
all the things where the microcode didn't get in your way.  MOVL2 would
have been about the same speed, but MOVZWL would have been slower.

> The 725 was faster?  I always thought the 725 was a stripped-down (and
> slower) version of the 730.  

Yes.

>                             (I remember decommissioning a 730, or
> maybe a 725, I forget...we were trying to find the CPU and eventually
> decided, only half in jest, that t'was a Z-80 running a VAX emulator.)

The 725/730's CPU was two entire boards full of discrete-looking chips, a
lot of which were ROMs and PALs.  The uVax1 was the first single-chip VAX
implementation, and it was a f---ing miracle for its time.