[Info-vax] Dave Cutler, Prism, DEC, Microsoft, etc.

[Neil Rieck]

>
> Gads.  Take this as a lesson: had they been able to produce nVAX on time,
> there wouldn't have been a need for Alpha.  From the blank-slate
> perspective, Alpha is great.  BUT it was a DISASTER for VMS regardless of
> the performance.  It broke binary compatibility, shedding ISV's and products
> that were never ported.  It caused calendar *years* of effort and millions
> and millions of dollars to do.  It was ported to VMS because at the time,
> VMS was still big business - but by then internally VMS was seen as a
> strategic dead end (at the time it was UNIX as the next big thing) and the
> UNIX/RISC weed had rooted.  VMS never recovered from the VAX running out of
> gas, and the port to Alpha.  But it was inevitable that it had to happen,
> and we (DEC) needed to have been planning what came *after* VMS.
>
> From the parochial viewpoint of the VMS installed base - what they needed
> was faster VAXes.  Not Alpha.
>
I'm not sure I agree with these points. In the book "DEC is Dead, Long
Live DEC" it was stated that clock-for-clock comparisons between CISC
and RISC showed that RISC out performed CISC two-to-one without any
special tuning. Also, research present to DEC management by Gordon
Bell stated that the computer industry was going through major changes
every approx every ten years so that the end of VAX was in sight.

Let's also remember that DEC/Compaq sold more Alphas than
VAX.Apparently Intel has sold more Itaniums than VAX + Alpha combined
so the world didn't exactly fall to pieces with the end of VAX.

What I really question is the speedup specs from Itanium (EPIC). Many
have said that the EPIC compilers never lived up to the hype, but that
Itanium is still twice as fast as Alpha. Well, we all know a system
will enjoy a huge speed increase just going from DDR memory to DDR2 so
maybe the biggest waste of money involved replacing Alpha (and PA-
RISC) with Itanium. And considering the Itanium delays, HPQ should
have followed EV7 with EV8. Intel should have followed x86 with XEON
versions of x86-64 and forgotten about Itanium. Anyway, Itanium is
with us now and we'll all have to migrate there or forget about any
future with OpenVMS (shudder)

NSR

p.s. This week I purchased an HP PC based upon the Intel Core-i7 which
contains CSI (common system interconnect) aka QPI (Quick Path
Interconnect) which was developed by DEC for Alpha. This technology,
which DEC had in 2002, took 6 years to make it into other Intel
products. Can you imagine what the CPU industry would be like today if
Curly + Carly hadn't killed Alpha?