RE: HP hostility to alpha holding back itanium?

From: Main, Kerry (kerry.main_at_hp.com)
Date: 05/05/05 

Date: Wed, 4 May 2005 20:32:48 -0400

> -----Original Message-----
> From: Bill Todd [mailto:billtodd@metrocast.net]
> Sent: May 4, 2005 5:40 PM
> To: Info-VAX@Mvb.Saic.Com
> Subject: Re: HP hostility to alpha holding back itanium?
>
> Main, Kerry wrote:
>
> ...
>
> > That same argument has been around since the x86 was first
> born (what
> > was it - 15, 20 years ago?)
>
> Speaking from ignorance as usual, I see. The 8086 was introduced in
> 1078, the 8088 (the slightly hobbled sibling around which IBM
> built its
> PC) in 1979, and the 80386 (the first 32-bit member of the
> family) in 1985.
>

Mmm... 1078 eh? Thx for the history lesson. I will be sure to remember
this.

:-)

>
> i.e. "this newest x86 is so cool there is
> > nothing it can't do, so everyone else should just move to
> it and forget
> > everything else."
>
> And speaking the usual horse*** as well.
>

Another bad hair day, eh?

Who cares what year or even what century the PC was born?

Geeezz - saying that one cpu architecture is going to force all others
out (as JF was insinuating with his reference to x86's replacing
mainframes) is ridiculous - no matter what or how good the cpu
architecture is.

Lets keep things in perspective here. Server performance and other HW
features is only one of many considerations when looking at new
platforms.

Remember that Customers buy computers to run applications to address
business requirements.

The server HW is important to Cust IT staff, vendors and newsgroup
participants, but it is only a small part of any IT infrastructure
costs. Staff, support and software costs typically dwarf the overall hw
costs and the server is only a sub-component of the HW (storage, network
etc) costs. Getting a Customer to replace or rewrite 10-15-30 years of
investment in business logic to another OS because that OS has a new
chip available that is x% faster than the Cust's current vendor chip
strategy "at a particular point in time" is just not in the cards.

The *costs and risks* to migrate, test (this is huge), QA and re-license
the business software, tools, development and operations utilities,
custom code, not to mention re-training or re-staffing (who have
technical skills, but no business logic specific to that environment) of
moving to an entirely different OS is the 800lb gorilla - not the server
HW.

And if you have talked to any CEO lately, they would tell you that they
want to *cut* IT costs and invest in areas that help their business
improve the bottom line. Any CIO that goes to the CEO and requests
buckets of $'s to port to any new OS platform better have an incredibly
good TCO and ROI justification that takes into consideration all of the
risks and costs I just mentioned. Either that or the CIO will be
pounding the pavement in short order.

What CEO's want today is not "rip and replace", but rather "integrate
new technologies and make better use of what we have".

>From an OpenVMS perspective, that is why the Itanium/Alpha/VAX mixed
clusters support is a very good investment protection strategy. And from
an obviously biased vendor viewpoint, this mixed cluster support
represents a brilliant OS architecture i.e. what other OS platform
supports 3 different architectures (CISC, RISC, EPIC) in the same
cluster running different OS versions and all accessing the same data?

[insert personal insults and spelling corrections from comp.os.vms.
moderator]

Regards

Kerry Main
Senior Consultant
HP Services Canada
Voice: 613-592-4660
Fax: 613-591-4477
kerryDOTmainAThpDOTcom
(remove the DOT's and AT)

"OpenVMS has always had integrity ..
Now, Integrity has OpenVMS .."