NetNews Usenet Archive 1992 #30/NN_1992_30.iso/spool/comp/sys/mac/hardware/24092

Internet Message Format | 1992-12-14 | 3.9 KB

Path: sparky!uunet!europa.asd.contel.com!emory!swrinde!zaphod.mps.ohio-state.edu!cs.utexas.edu!sun-barr!ames!data.nas.nasa.gov!taligent!apple!goofy!mumbo.apple.com!times!sirius.aux.apple.com!coolidge
From: [e-mail] (John L. Coolidge)
Newsgroups: comp.sys.mac.hardware
Subject: Re: P5 v. PowerPC (WAS: Where the mac really wins)
Message-ID: <[e-mail]>
Date: 14 Dec 92 23:46:04 GMT
References: <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <-ZR=[e-mail]>
Sender: [e-mail] (News Subsystem)
Reply-To: [e-mail]
Organization: Open Systems Development, Porting&I/O Group, Apple Computer, Inc.,Summary:
Lines: 60

[e-mail] (Chung Hsiung Yang) writes:
>In article <[e-mail]>, [e-mail] (Jan-Olov Lantto) writes:
>> In article <[e-mail]> [e-mail] (David Wagner) writes:
>> There are some big differences between PowerPC and Alpha, that makes Alpha
>> unsuitible as a chip for PC s.
>>
>> 1.  Chip area. Smaller chip area = lower production cost = cheap computers.
>> 2   Power consumption. Low power consumption = can be used in notebooks.

>    DEC Alpha's are is made with sub-micrometer technology.  I believe
>it is something like 0.7 micron technology, which allows the chip voltage
>voltage to be 0 - 3.3 V (compared to 0-5v for most of the PC chips today).
>[...]

The real killer for the current generation of Alpha chips for use in
PC's is price. The current Alpha implementation is _very_ expensive
(I've seen claims that it's well over >1,000/chip, maybe over
2,500/chip). These prices are fine for workstations which list at
$14,995 minimum, but they're completely unsuitable for PC's which
need to be able to hit a 2-3,000/system price point. Looking over what
goes into that $14,995 workstation, it seems not at all unlikely that
the chips are in fact costing DEC a couple thousand dollars each.

The main problem that I see with current Alpha implementations is that
it's just a bit leading-edge. Sure, you can design a really nice chip
set if you run with the very latest in design and fab technology (as
I'm told they have; not being a chip hardware person, I don't know
myself). The problem is that then you're stuck running with
leading-edge stuff (which is costly) every time you scale the family.
If you don't, you fall back to where your competition is on
performance.

Note, too, that I'm not commenting on the Alpha _architecture_, which
at least theoretically could be implementation in much cheaper (and
slower) silicon. I'm only commenting on the current chips.

>    For notebooks, you could include power management, by varing the
>speed of the processor - to slow down the clock speed when the processor
>sits idle.  But using DEC Alpha in notebooks would be an overkill, I agree.

Nah. There's a definite market for 60+ SPECMark notebooks. I think
you'll see them mass-marketed in a couple years. There are SPARC
notebooks now. I think you'll see PowerPC notebooks fairly quickly
after desktop and deskside models debut (indeed, it's now reported
that IBM wants to put out a PowerPC 601 notebook in very early 1994
running AIX); that's a 40+ SPECMark system.

Probably you'll see Alpha notebooks down the road, but they probably
won't be based on the current chips. They're way too power-hungry when
run fast, and they're way too expensive to be run slow.

--John

        Thanks to the diligence of the FBI, this particular vacuum
        cleaner won't fall into enemy hands.
                -- Howard Hughes, _The_Rocketeer_

+++John L. [e-mail]+++++++++++++++++++++++
I speak for myself, not for Apple Computer. Copyright 1992 John L. Coolidge.
Copying allowed only if attributed, and if all copies may be further copied.