Conference 7.286::digital

    
    would be a great time to switch from "Whatever it takes" to
    
    "We have it NOW!"
    
    

    
    
    Back to the old one your saying? "Digital has it now."
    

RE: .0

	Kerry, you've posted this in at least 4 conferences. Could
	you not just post pointers to one single place to discuss
	this? (Or at least SAY that you've posted it in many 
	places?)

						mike

Yeah Kerry

Get your act together eh?

:-)

  Kerry,
  
  Thanks for posting this here and anywhere else that is relevant.
  Our own marketing organization will get this out later when its
  old news!  We need all the boosts to our story we can read.  So what if
  we see it more than once!  But it might be useful to indicate the
  discussion thread should be in conference X.
  
  keep up the great work,
  regards,
  -ian

> So what if we see it more than once!

  Just this; with it posted in several conferences, some of which I do not
  read, I might miss some relevant discussion or news about the topic.  It
  is difficult to follow a single string of notes as it is. 8^)

  	mikeP

    Is P7 the joint Hp/Intel vapour-chip, or does that come after p7?
    
    Just trying to make sure I spread correct FUD around migrations ;^)

    The following arrived today.  My question is around 64 bit NT.  It
    seems to say that MS and Intel will synchronize the delivery of the
    products.  
    
    We should be able to buy another year or so with the only 64 bit NT
    available if Intel's chip slips and 64 bit NT will indeed be ready on
    Alpha.  Of course Intel has no reason to want to even mention that
    other 64 bit NT's might be ready, but why isn't MS saying so?
    
    Again, no mention whatsoever of Digital and our efforts around 64 bit. 
    What gives?
    
    SANTA CLARA, Calif.--Intel Corp.'s first 64-bit processor, code-named
    Merced, is now scheduled to hit the market sometime in 1999.
    
    Contrary to widely published reports that late 1998 was the target date
    for the processor, Intel spokeswoman Marion Koehler said the company has
    never publicly committed to a time frame except to say the chip would be
    released by the year 2000.
    
    >>The shift means that Microsoft Corp.'s 64-bit version of Windows NT
    >>won't appear until 1999. Microsoft and Intel executives said at last
    >>September's NetWorld+Interop show that they would synchronize the release 
    >>of their respective 64-bit products.
    
    Merced is a high-end server processor with both CISC and RISC
    properties. Intel worked with Hewlett-Packard Co. on the initial design; 
    however,Intel will be responsible for manufacturing, marketing and 
    licensing.
    
    <<PC Week -- 02-20-97>>
    
    

>> So what if we see it more than once!
>   Just this; with it posted in several conferences, some of which I do not
>   read, I might miss some relevant discussion or news about the topic.  It
>   is difficult to follow a single string of notes as it is. 8^)

	The old kitchen sink defense.  Also known as "let's merge
	all notesfiles into one".  Ya, that will work real well ...

re .8: see note 3157. Do not confuse "64 bit NT" (which may well be 
delayed until P7 comes out) and "64 bit VLM support on NT" (which 
apparently will be in NT 5.0 for Alpha only, end of this CY/beginning 
of next).

	Vincent

>re .8: see note 3157. Do not confuse "64 bit NT" (which may well be 
>delayed until P7 comes out) and "64 bit VLM support on NT" (which 
>apparently will be in NT 5.0 for Alpha only, end of this CY/beginning 
>of next).

don't tell us, tell the rest of the world!

that's directed at MARKETING, not the author of the above comment  8-)

-mark

    No matter what they call it, if SQL Server wants to compete with Oracle
    in the high end / corporate database market, M$ has to build 64 Bit 
    support into NT quickly.
    
    And the ONLY processor they currently can implement it on is Alpha. 
    
    Klaus

    RE: .11, .12
    
    Exactly!
    
    We should be all over this announcement from a marketing standpoint.
    
    Tell the world that Intel can't really figure out how to make a 64 bit
    chip, and by the way, we did it years ago.  We also have a 64 bit
    operating system and many many 64 bit optimized applications.
    
    NT will be 64 bit on Digital year(s) before Intel even is able to
    produce a 64 bit chip.
    
    But then again, maybe we don't want to hurt Intel's feelings.  
    
    Ken
                                                    
    Intel IS NOT the only alternative for running NT.  
    

    .13
    
    
    let us read Art of War before we go aggressively marketing against
    Intel per se.
    
    
     

Yes, we should avoid 'aggressively' marketing _against_ Intel, but we could be 
blowing our own horn real loud right now.

    Before we go trumpeting our 64 bit benefits to the world at large, 
    the marketplace needs to undergo a paradigm shift. Except for a few 
    niche areas like Very Large Databases, most users don't feel any 
    real need for 64 bits.
    
    Much of their current software is 16 or 32 bits, and little is pushing
    the envelope on 32 bit address space. A key operating system, Windows
    NT, does not support 64 bit operation and the timetable for conversion
    is in question (see other recent notes this conference). 
    
    We're like the early 1960's ball point pen manufacturers who introduced 
    their products by advertising that they could "write underwater". Not many
    users had occasion to do underwater writing. Fountain pens were proven
    technology.
    
    Of course, eventually the paradigm will shift. Ballpoint pens proved
    cheaper to manufacture than fountain pens and were much more convenient
    to use. Similarly new application types (perhaps internet servers?)
    will force the move to 64 bits.
    
    Until then, I'm not sure how well the 64 bit theme plays. Our
    competitors have much room to create and distribute new FUD on this
    topic. I believe we have to help create a market paradigm shift first.
    
    This is part of the joy of being early to market with a new concept.  
         

    Also, apps running under 64 bit NT require that the physical memory
    actually be there (no page fault mechanism), so the box must have
    >2Gb of memory installed AND have an application that uses the 64
    bit API's.  Not that many of the current or past Alpha models can
    even go over 2Gb of memory installed, so they won't see any benefit.
    On the other hand, perhaps a year from now 2Gb will be considered
    bare bones.
    .02 K

    yeah, but not alot of the retail PC world needs MMX or 200MHz; yet you
    ask any home-computer wonk, and these are must-haves.. 
    
    The same happens at all levels of computer buying.
    
    Waiting till the world believes it needs 64-bit is "sales"; convincing
    the world it needs it *now* is "marketing".
    
    I'll let everybody decide for themselves which goes on here.
    
    ...tom

yes, and come to think of it, I don't recall _needing_ 32 bits...  did I?

regardless, if its important enough for Intel to say when they'll be ready, it
should be important enough for us to say we have it now (or sooner or whatever).

As for writing with ball point pens underwater, doesn't the paper get soggy? 

    re: Note 5157.19 by 12680::MCCUSKER
    
    > yes, and come to think of it, I don't recall _needing_ 32 bits...
    
    You must not be a software engineer or application developer.
    
    Ken....

    RE: .19
    
    I certainly remember.  While I don't know if I've ever used the 32nd
    bit, I did enough overlays on RSX-11M to be happy when I saw those
    18,19,20,21,22,etc. bits when I first saw a VAX.
    
    				-John

    Re last two....
    
    Yeah, yeah, yeah... from an engineers perspective there was a need.  
    I was speaking from the end users point of view.  Sorry I said it.

RE: .22

End users see the benefits of 32-bit addressing on PCs in lots of 
subtle ways.  Like applications that run more quickly or don't crash 
as often, or can support big documents.

--PSW

RE: .7

As I understand it, there are two P7 efforts underway at Intel.  One 
is the P7 replacement for the P6, and the other is the joint venture 
with HP.  Both are farther away than they appear.

--PSW

re .24

	>> Both are farther away than they appear.

The question is, will we let this window of opportunity
pass?  I think in our hearts, we all know the answer.

    Does Nintendo need 64 bits?
    Who cares, its a great hook to hang your hat on.
    
    If we look at the Kano model, we have 3 types of product element:
    Dissatisfiers 	- "must-be", "basic", "expected"
    Satisfiers		- "wanted, and usually asked for"
    Delighters		-"attractive" or "exciting" characterisitcs
    
    Customers will only tell you about the satisfiers. The things that they 
    understand and can articulate. They wont tell you about the
    dissatisfiers, because they expect those things to be there and they
    expect you to know about them and what is expected of them. They wont
    tel you about the Delighters, because they dont have them in their
    minds when they tell you what they want.
    
    For instance, we pitch 64 bit computing as a delighter. The "out of the
    box" surprise that will make you love Digital. Unfortunately, if you
    dont meet the other elements required by the customer, there is NO
    point in having the 64 bit capability.
    
    For instance:
    A customer may have - "good manageability" as a Dissatisfier. They wont
    tell you this, they just expect your system to be easy to manage.
    They may tell you they require "support for Internet protocols"
    
    If you fail to deliver on these 2, your gone. It doesnt matter, for
    the vast majority of customers, that you have 64 bit computing, You
    havent met the fundamentals.
    
    However, if you meet the fundamentals, AND deliver them extra delight
    by 64 bits, then you win big. BUT, it is not a substitute to meeting
    the fundamentals. Give great 64 bit and lousy system management or
    internet connectivity, you lose.
    
    OK, so some possible dissatisfiers and satisfiers, before 64 bits:
    
    . Reliable delivery channels
    . Sensible pricing models that suit the customer
    . Many choices of application availability
    . General market confidence in company/product
    . Knowledgable,  motivated front people to the organisation
    . Seen to be a high quality leader
    . Able to help my business to the next big business success, a technology
      leader
    . Consistency
    . Clear articulation of purpose and place in the world
    . Good ongoing customer relationships
    . Affinity for what me as a customer or potential customer is trying to
      achieve
    . Obvious ways for me to communicate back through the company on
      various matters.
    
    Some of these are unwritten, you wont see them in a tender or even see
    them mentioned by anyone. They are expected of us.
    
    There is nothing wrong with 64 bit computing or 1-3-9 for that matter.
    But where they  miss is in the lack of focus they give us in meeting the 
    fundamentals that business requires of Digital and Digital requires of
    its business.
    
    Concurrent with the product strategies, 1-3-9 and Alpha 64 bit yeah
    team, we need a broader unifying strategy that attempts to capture and
    address ALL of the perceived things we fall down on and do right that our 
    customers perceive as fundamental. Until we do this, any product strategy 
    will produce sub-optimal results. Our continued floundering attempts at
    re-engineering, "Supply chain and ball", "Hug your subordinate day",
    "Company values, we love 'em"...
    are all hollow and will fail unless they are built around a context of
    the end goal we are seeking for ourselves. My pass at it is:
    
    "Profitable market acknowledged high quality goods and services 
    with no impediment to customer purchase from a recognised high quality
    company that provides no impediment to doing business and communicates,
    via the media and via enthusistic motivated employees, with its
    customers."
    
    I believe this statement shows where we want to be, in the eye of the
    customer, and highlights many of the deficiencies the company currently
    faces that must be addressed for this company to get the leverage we
    need out of our product positioning.
    
    - John
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    

    re: .25 and window of opportunity
    
    you mean like the Rainbow ? - able to run the old (CP/M) and the new
    (CP/M86 and MS-DOS) transparently on the same box. Or the desktop
    'personal' VAX - which never made it to market :'( the closest was the
    VLC. Or who knows what other wonders have been still-born in this
    company?
    
    H

    Just an example, who needs 64 Bits (from an end-user perspective):
    
    In my area (manufacturing industry / product engineering), a core
    application is Product Data Management (PDM). This market is growing at 
    30%+ per year and many of our large customers have PDM projects under
    way. It's the kind of customer we used to sell VMS clusters to, btw.
    
    PDM installations often require a million or more documents to be
    handled. The central index for this database is easily larger than two
    GB and is accessed read-only in most cases. It's a matter of large
    database performance and benefits from VLM64.
    
    So, what's the scoop ? If our competitors bid for these server
    installations, quite often they will bid two machines: An application
    server plus a database server. What we can do: Add a lot of memory and
    win the project with just one single server running both the
    application and the database.
    
    There is no need for a religious debate about 32 vs 64 Bit or about
    VLM. It's simply price and price/performance and it sells.
    
    M$ cannot adress this market at all today, it's plain UNIX at the
    moment. This is despite the fact that a number of customers would like
    to use NT - they just cant't do it. Unless M$ has a VLM64 database
    product on NT.
    
    Klaus
    
    

    Linux users are all rarin' for 64-bit technology.  Just check the ads.
    
    --Dominic

	If it can't be done with 12 bits, it doesn't need to be done!! 
a flash from the past.


Keith

    No, I think that 36 bits is the criteria!

>               <<< Note 5157.16 by MSDOA::GUIDRY "Ghost Rider" >>>

>    Until then, I'm not sure how well the 64 bit theme plays. Our
>    competitors have much room to create and distribute new FUD on this
>    topic. I believe we have to help create a market paradigm shift first.

And our competitors apparently have some powerful new FUD in a Gartner Group
Report that we're told dumps on DIGITAL UNIX. Internally, a rebuttal was
circulated widely on Friday, but not the Gartner report itself. [The rebuttal
can't be posted because it's marked "DIGITAL CONFIDENTIAL".]

    
    Re .31:
    
    I still have the orange T-shirt: "If your computer doesn't have 36
    bits, you're not playing with a full DEC."
    

    Even if your customer is convinced s/he has no need of 64 or 33 bits,
    s/he will have to migrate to 64-bits at some point.  Even if the
    migration requires no programming or tuning changes (right), s/he will
    still have to go through a rigorous, expensive QA.  Why not start with
    64 and not worry about the migration.
    
    Customers who see no benefit from 64-bits are usually the easiest to
    scare.
    

    .3 (Mike)
    
    Apologies if my posting the .0 note in a few notesfiles ruffled a few
    feathers, :-) but there are numerous folks who follow only a few
    conferences, so it is very easy for them to miss postings of this
    nature. As an example, in addition to notesfiles, a number of people
    also follow CompuServe, newsgroups, online mag's etc.
    
    Anyway. I like the suggestion of recommending what conference the
    discussion should take place in - will do that in future when cross
    posting info ..
    
    :-)
    
    Regards,
    
    / Kerry

http://www.winntmag.com/

    533 + 64 = ONE (HECK OF A SOLUTION) 

As reported in last week's Windows NT Magazine UPDATE, Microsoft said
it will not wait for the the upcoming P7 (Merced) processor (due in 1999) to
begin shipping a 64-bit version of NT 5.0. We also reported that the first
64-bit version of NT 5.0 would run on Digital Equipment's (DEC) Alpha
processors. Since then, we've heard some rumors that Microsoft is buying
up quite a few of the new DEC Alpha 500MHz machines for its development
staff. And now that DEC is directly involved in porting Microsoft desktop
applications to the Alpha platform, non-developers at Microsoft are getting in
on the new Alpha fun. 

All the pieces are in place for DEC to mount a full assault on the NT
marketplace. Digital announced this week that it will slash prices on new
Alpha systems, creating platforms for under $3000, and Digital also showed
that its newest Alpha chips clock in at a whopping 533MHz. Microsoft's
decision not to wait on Intel is a huge break for Digital; Digital now can have
a two year head start on Intel by providing the only platform for 64-bit NT. Not
only will Intel feel the pressure because of this decision, but so will UNIX
vendors. Surveys show that many UNIX shops are eagerly awaiting a 64-bit
NT platform, and with Microsoft's upcoming Wolfpack technology (which will
begin to show up in Service Pack 3 for NT 4.0), this opportunity is the right
high-end combination the UNIX and NT world have been waiting for. 

 

                                    Mark Joseph Edwards