I am being slammed by email today so I will have to be brief. I was hired in 1987 by Mizar, Incorporated, to bring up AT&T SVR4 v2 Unix (sic) for them to sell as a VMEBus product. IIRC my development machine at that time had a 321SB board in it. My understanding is that the only difference between it and the 3B2 machines was the bus interface. VMEBus was going to be "the future" of expansion buses for industrial computers at that time.
Even by the standards of the time this CPU was underpowered. IIRC we made our own boards based on it. (VMEBus master boards made by manufacturers who did not specialize in VMEBus, as we did, tended not to be well-designed, and only implemented a minimum of features. Their multiple bus master support was non-existant, for instance.) The product did not sell. After a year they switched me over to the design of the next "future" processor, a sparc32 board. That did not sell well, either. Then we tried to sell the idea of MIPS-based boards to interested customers but not much happened. During that time Mizar was making most of its money off the 68k product lines. The 68000-, 68010-, and 68020- VMEBus machines were still selling well. We were selling these machines either to run bare-metal, PDOS, or vxWorks. (The latter was what I was assigned to the whole time.) Older people will recall that Motorola "lost its mojo" around that time and were having trouble scaling up the 68040 to faster clock speeds (this is because much of their technical design team walked out the door to start Ross Technology.) So we were trying to figure out what we could offer to stay competitive. tl;dr: we didn't. I left at the end of 1989 after running my mouth to management too much. I can go on a treasure-hunt to see what I still have from those days but it won't be this week. I do have manuals and maybe even a board or two. AFAIK no one else ever even got AT&T-based processors out the door at all. mcl
