[Info-vax] RealWorldTech on Poulson

[Michael Kraemer]

JF Mezei schrieb:
> 1 To state that DEC did try to pitch Alpha to the PC market may be
> technically true, but only technically.
> 
> To penetrate a market, you need to have staying power and invest in
> marketing. You don't just introduce a product and if it doesn't pan out
> within 6 months declare that attempt failed and pull out.
> 
> DEC pulled out of the PC market before Alpha was given a REAL chance of
> succeeding.

DEC tried as long as they lasted,
in the beginning with the Jensen, towards
the end with the "Vobis" Alpha PC.
But the PC market has its own rules, if a product
does not succeed within a few months, it will
be dumped.

> 
> Look at Microsoft. It has staying power and despite all its failures in
> the mobile market it is still going at it, and with each iteration, it
> mamages to win some sideline market (for instance: Ford Sync,  TV  set
> top controllers which also brought in sales for the entire IPTV product
> line (Media Center).

Because their product fits customer demand.
Alpha PCs didn't. The only advantage
the Alpha could bring was higher raw performance,
but at a price. With $2000+ per CPU you can't
build a $1000 PC.
But even that wouldn't have helped,
considering the two main reasons people
bought PCs in the 1990s: Internet and games.
I don't see how a 64kbit connection would run
faster if you have Alpha instead of intel inside.
And just how many games were offered in native Alpha versions?

> 2- It may be true that the initial iterations of Alpha were not
> impressive. 

Since you have exactly one chance to make a
first impression, it is of utmost importance.
A mere 30% performance (and almost no price/performance) advantage
over the competition is not instrumental in winning customers who
have long term investments in competitor's platforms.

> But that is because they focused first on building a very
> strong foundation that was clean and would let the architecture grow
> over at least 2 decades.

This would only have panned out in an ideal world,
i.e. without competition. With the introduction as late
as 1993, there simply was no time left.
Had DEC put one of their three or four paper RISCs
into silicon already in the late 1980's, they would have
had a chance. Likewise, if they had stayed with their Mips
decision.
But starting from scratch with Alpha and burning money,
when all the competition was already earning money,
that was really gaga.