Still OT: Pentium / M$ drivel... (fairly long)

From: Roger Merchberger <>
Date: Tue Dec 29 20:23:21 1998

Once upon a midnight dreary, Tim Hotze had spoken clearly:

>A Pentium as in a Pentium or a PII? If I Understand Correctly
>(IIUC), it's just a PII with no external casing, and either no cache
>or 128K cache. Even SMP is supported. (Dual Celeron 333...
>nice.... cheap, for a 333MHz SMP system...)

Noper... out of the box the Celery processor does *not* support SMP. Intel
*deliberately* disabled it in the design of the processor... there's a
particular pin on the Celery that says: "Nope. Sorry. Can only have one of
me." Now, if you do a websearch for prolly "+dual +celeron +modification
+SMP" on altavista then you should come up with a page (on a Japanese
server) that a person found out which traces to cut, and where to solder
wires to re-make the SMP connections to run duals... but you're (obviously)
going to void the warranty on the CPU's, etc., etc., etc.

>> The dynamics of the microprocessor market is more complex than you think.
>> If MIPS was so good it would ahve pushed out x86. Alpha is a 64bit cpu
>> targetted at high end systems and the MicroVAX (32bit) was already well
>> established and faster than 386/486.

I have to disagree with this, Allison, one of my points being: The value /
speed / whatever of a Microprocessor means little versus the *marketing* of
a processor / computer. The lowly Motorola 6809 at 2 Mhz outperforms a
10Mhz 80286... by far. It also *smokes* the 6502. And other than Xenix 286
(super-expensive), OS-9 is the most powerful OS available for these
processors, and at a reasonable cost (When I bought my copy - $139.)...

But how many Tandy CoCo's (1, 2, or 3) were sold in relation to Commie
64's, Atari 800's, and XT/AT's??? (Never mind that the CoCo had the
longest-running continuous production for an 8-bit home machine)... The
CoCo came in last in this particular list.

Marketing - not capability. MIPS didn't have a chance to "take over" in the
main desktop market... just be a niche market deal - like the Sparc's,
Alpha, etc.

> But MIPS are becoming common... they power the fastest of
> Windows CE HP/C's, the ultra-fast N64 (which, BTW, only has
> 4MB RAM but can easily outperform any PII 450 with a 90MHz
> video card and 128MB RAM, if you look at realtime 3D).

The N64 does better 3D - because it's designed for *solely* that. Can it
run a spreadsheet? Can it run a WP? Hell, even a terminal proggie? The MIPS
means little in this - the graphics subsystem chipsets are where you get
your 3D speed. You're comparing apples to oranges there. A Maserati is
designed to go a *lot* faster than my 4-wheel-drive - but I can run over a
lot of Maserati's with my truck. (and with a top cruising speed of 115mph,
it's no slouch... ;^> )

The reason MIPS are becoming more common is because of *marketing*, not
technical savvy. My Palm III beats the tar out of any WinBlows CE machine -
Let's compare specs:

Palm III WinBlows CE
16-Mhz 68000(dragonball) 50 / 75-Mhz MIPS (that I've seen, anyway...)
800K OS size 8 Meg OS Size (except one that made it in 4 Meg)
2Meg RAM 4 Meg RAM
1.95Meg Usable RAM 1.5Meg Usable RAM
20-80K avg. pgm. size 120-300+K avg. pgm. size
~90 hours cont. use
   on 2 AAA Batteries Giggle. Smirk. Gasp. Battery Life?!?!?!?!?

Now, the Palm III is outselling any *individual* WinCE platform - but by
sheer numbers of companies giving (IMNSHO) way too much money to Billy-bob
Gates, the Palm is not the market leader. Flooding the market certainly has
it's advantages.

Now, to bring this (slightly) on-topic - all those assembly language skills
you learned *way back* on one of those nice, *fast* classic machines, like
an Atari ST, or early Macintosh, or Amiga (damn... *still* waiting for my
1200's) - fit very nicely to the Palm realm. What can you do on a WinBlows
machine? Not much - Billy-boy doesn't release all the OS / API information
(at least freely) so you can't program in Assy. or another
"non-M$-approved" programming language. --- Well, not easily.

>> Allison

Roger "Merch" Merchberger
Received on Tue Dec 29 1998 - 20:23:21 GMT

This archive was generated by hypermail 2.3.0 : Fri Oct 10 2014 - 23:30:51 BST