[Info-vax] Vaxes shutting off this week

[johnwallace4 at yahoo.co.uk]

On Mar 1, 7:38 pm, JF Mezei <jfmezei.spam... at vaxination.ca> wrote:
> Michael Moroney wrote:
> > A big difference is that the PDP and the VAX ran their courses and were
> > discontinued at an appropiate time.  Alpha was killed when it was pretty
> > much the fastest processor.
>
> Upper management were convinced to sponsor multiple projects. One was
> the 9000, the other was the Nvax, and there was the skunk work Alpha
> project.
>
> 9000 tanked. The nVAX ended up succesful, but the Alpha team came up
> with Alpha and then DEC stopped developping the VAX.
>
> So, now, lets revion history (one of the few times we can do this)
>
> The 8086 was considered dead-end, so Intel started what would be IA64.
> IA64 turned out to be a dud that was over half a decade too late.
> Meanwhile, Intel was forced to keep upgrading the 8086 because IA64
> wasn't ready, and with the Alpha-inspiration, the pentium 3 turned out
> to be really fast and once the barrier was broken, the 8086 achieved
> incredible growth in performance, making the IA64 irrelevant since not
> only had the prediced 8086 barrier been broken, but the 8086 was giving
> extremely respectable performance.
>
> Had Intel released Merced in 1995 and a not-laughable IA64 in 1996-1997,
> it is possible that IA64 might have totally replaced the 8086 as the
> industry standard chip. By the time it arrived, IA64 was useless since
> the 8086 had already made IA64 a proprietary niche product.
>
> VAX was never really allowed to break its barrier because Alpha came too
> quickly to market.
>
> With 25-JUN-2001 as end of architecture no matter what, it is likely
> that VAX upgrades could have kept up with the Jones' until then.
> However, Alpha had been meant to have a 25 year lifespan. And for such a
> long life, it woudl be quite likely that VAX would have runned out of
> steam well before and Alpha would have then sped past VAX
>  for the remainder of the 25 years.
>
> So, for a 25 year timeline, Alpha was the right choice beause it offered
>  growth for the whole period. But for a 10 year history, it would have
> been cheaper to just upgrade VAX until 25-JUN-2001 and forget about the
> port to ALpha and the abandonment of so much software.

I don't see any mention of the influence of the AMD64 project, but I
know that if there'd been no AMD64 there'd probably have been no Intel-
badged AMD64 clone.

If/when an x86-64 has the same dozens of MB of on-chip cache as
Itanium currently has, which one would be the winner in performance,
and in manufacturing costs? (Hint: manufacturing costs can be quite
dependent on volumes, low volumes are A Bad Thing).