Re: Next Gen Fabs & Itanium
From: Robert Deininger (rdeininger_at_mindspringdot.com)
Date: 01/19/05
- Next message: JF Mezei: "Re: carly(tm) transcipt"
- Previous message: johnhreinhardt_at_yahoo.com: "Re: Error log message"
- In reply to: Dave Froble: "Re: Next Gen Fabs & Itanium"
- Next in thread: JF Mezei: "Re: Next Gen Fabs & Itanium"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Wed, 19 Jan 2005 04:33:36 GMT
In article <41ED5C5E.1070200@tsoft-inc.com>, Dave Froble
<davef@tsoft-inc.com> wrote:
>Robert Deininger wrote:
>
>> In article <41EC90D7.2020005@tsoft-inc.com>, Dave Froble
>> <davef@tsoft-inc.com> wrote:
>>
>>
>>>Keith Parris wrote:
>>>
>>>
>>>>Integrity Servers cost less than the equivalent Alpha box (thanks to
>>>>economies of scale). So it's not hard to convince intelligent people to
>>>>purchase them.
>>>>
>>>
>>>Can you be specific about "economies of scale"?
>>>
>>>
>>
>> ...
>>
>>
>>>Does an Alpha box really cost more to mfg? My bet is that HP is charging a
>>>premium for Alpha, since the only customers left don't have any other
choices.
>>>A bit like a MicroVAX 3100 model 98 costing twice as much as an Alpha
that was
>>>10 times faster.
>>>
>>
>> Alpha systems absolutely cost more to make than similar Integrity
>> systems. You'd lose your bet if the numbers were available to the
>> public. Since they're not, you're safe with your wild public
>> speculations.
>
>
>Nice shot, but is doesn't address the question.
See my reply to J.F.
>> The HP system designers are have traditionally been much more cost-aware
>> than the Alpha system designers at Digital. Integrity systems have much
>> more re-use of components and subsystems, for example. The HP design
>> teams are just better at low-cost design than their Digital counterparts,
>> likely because their managers have pushed for low cost for years. There
>> is a cultural difference that couldn't happen overnight even if Digital
>> had wanted to design lower-cost servers.
>>
>
>Your responce seems to indicate that it's the system design, not the CPU chip
>that makes the difference. Does that indicate that the HP design teams could
>produce an Alpha CPU based system at a cost comparable to an IA-64 based
system?
Yes, the whole system design determines the cost. The CPU chip is just
one part of the problem.
I suspect either team could come up with a lower-cost Alpha server design,
but only if they got to start from scratch and not be tied to existing
design tradition. And only if low cost was a clear goal backed by
management. In that situation (which hasn't existed for Alpha for a
decade, maybe never) I don't know which team would do better.
Of course, starting from scratch would add a couple of years to the design
cycle, which might lead to a product with other problems.
DS15 was brought to market at breakneck speed. They had to re-use
(conceptually) almost everything, mainly from DS25 and DS10. They HAD to
finish quickly, but a secondary goal was to reduce the cost. They tried,
but the result is pretty disappointing compared to an rx1600 or rx2600.
If DS15 had abandoned the short design cycle and the reuse, it could have
been cheaper. But a DS15 12 or 18 or 24 months later might have been too
weak a product to be interesting in the marketplace.
>If not, then what makes a system using an Alpha CPU more expensive.
- Next message: JF Mezei: "Re: carly(tm) transcipt"
- Previous message: johnhreinhardt_at_yahoo.com: "Re: Error log message"
- In reply to: Dave Froble: "Re: Next Gen Fabs & Itanium"
- Next in thread: JF Mezei: "Re: Next Gen Fabs & Itanium"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|