[Info-vax] BOINC for VMS

[JF Mezei]

David Froble wrote:

> Well, the way I remember it, Ken wanted to extend the VAX, not create a new RISC 
> processor.  Regardless, on his watch the Alpha got it's start.

 Olsen had chosen the VAX 9000 with ECL tech as the future and canned
the other options which included new architecture (Cutler's pet project)
and the nVAX.

nVAX was close enough to completion that they finished it.

When the 9000 failed, Olsen reluctantly accepted to resurrect "plan B"
which became Alpha and formed a new group from the ashes left from when
Cutler left.

nVAX ended up outperforming the 9000 in both cost and performance, so it
became the architecture/tech that was used for the remaining VAX
products and was so good that VAXes remained viable for many years. Had
they spent more money developping it, I am sure they could have made
major performance improcements to it. But as a sunset architecture, it
just got a couple of  speed boosts once Alpha came out.


So the Alpha was a plan B for Olsen, and it was delayed not only because
 the project was "suspended" once the 9000 was declared to be the
direction for Digital, but also because once this happened, the team
that worked under Cutler see a loss of key personel.

Hindsight is 20-20 of course.  At the time, it may not have been obvious
that ECL was a costly dead end and that 64 bit was the future.


Had the nVax been chosen as the primary strategy for DEC instead of the
9000, it is quite possible that VAX would have continued with success
and been expanded to 64 bits once the market was ready for more RAM than
4gigs.


All water under the bridge now. History.