🚀 go-pugleaf

In article <1992Oct31.234604.21078@tamsun.tamu.edu> daugher@cs.tamu.edu writes:
>Well, rumor has it that IBM, Sun, HP, and DEC will all be announcing new
>RISC workstations November 10th or so.  The DEC unit will supposedly do
>200 MIPS on a single 200 MHz Alpha CPU (twice the MIPS of an IBM 3090/600J
>mainframe!).  Assuming it comes in under $20,000 it would hit NeXT's target
>ratio of $100 per MIP.
>
>If so, NeXT will have to respond promptly with some announcements of their
>own to avoid the appearance of being behind in development.  Two or four
>50 MHz 88110's should compete nicely; we'll have to wait and see if that's
>what NeXT actually does.

I'm actually quite worried about Nov. 10th. One of the machines Sun will
release is a Tsunami (sp?) based SPARC which delivers 28 SpecMarks for
less thatn $6k. Even though OpenWindows SUCKS, it's hard to justify
NeXT's 16 Specmark, no bus system for $8k.

The new HP will offer even better MIPS/$.

My impression of NeXT is that they aren't even close to releasing any
type of RISC machine. I believe that the vast majority of their efforts
have gone into NS486. It seems that NeXT is targeting the PC market
rather than take on the Workstation market. Granted, the line between
is rather blurred, but if NeXT is going to compete with Sun, SGI and HP 
they better get a RISC box fast.

Is NeXT trying to be the NeXT Microsoft?

As far as Intel CPUs go, aren't they the ones that expect mulitiple-byte
data to have bytes swaped? - Yuck.

Eric

This week's Info World, "Reports from the Field" claims that NeXT has
terminated the dual 88110 NRW and will go with the P5 (Pentium) from
Intel in future machines.

This sucks for a few reasons:
1. We won't have a faster NeXT till the P5 ships.
2. We're stuck with Intel.
3. The NRW will be AS FAST AS A HIGH-END CLONE.

Basically, if you stick with NeXT you commit to NeXTSTEP 486.
I was going to order a NeXTstaiton Turbo Color, but now it seems
that a BUC (Butt-Ugly-Clone) is the best hardware investment for
NeXTSTEP.

Hey, IRIX ain't so bad...maybe I should go for an R4000 machine after all.

Eric

In article <1992Nov2.170741.2092@jhunix.hcf.jhu.edu> eboltz@jhunix.hcf.jhu.edu (Eric Scott Boltz) writes:
>This week's Info World, "Reports from the Field" claims that NeXT has
>terminated the dual 88110 NRW and will go with the P5 (Pentium) from
>Intel in future machines.

Is this so suprising?  One of the things that stuck in my mind the most when
reading all the matterial I received on NeXTstep 486 was that it was tuned
to run fast on the 486 ( not 386 ), and when the the Pentium came out it would
be tuned for that chip using its unique features.  Hell, they didn't even do
that for the '040.  Seems to me it's just a matter of making it official and
SAYING they intend to go with the Intel chips ( but keep in mind they still
haven't ).

>This sucks for a few reasons:
>1. We won't have a faster NeXT till the P5 ships.

I'd be surprised if the Pentium doesn't ship in quantity before the 88110

>2. We're stuck with Intel.

Is that so bad?  More machines, ability to run X million DOS and Windoze
programs.  I think I'll survive.

>3. The NRW will be AS FAST AS A HIGH-END CLONE.

After two years of hearing about the mythical NRW that NeXT is working on,
the best they've come up with is uping the clock on the '040.  I'm tired
of waiting.  Those high-end clones are going to be sporting Pentiums by
next Spring.  Real hardware, real fast.

Sure, the Pentium isn't the best chip they could chose ( is it true they
STILL only have 8 registers? ), but I'll be happy to see them move forward
with ANYTHING.

>Basically, if you stick with NeXT you commit to NeXTSTEP 486.

Remember, NeXT has promised to update NeXTStep to be tuned for the Pentium.
I wish I had that on paper signed by SJ :-)

Chris

-- 
Christopher D. Audley
Elec & Comp Engineering
The Johns Hopkins University
Internet: audley@jhunix.hcf.jhu.edu

eboltz@jhunix.hcf.jhu.edu (Eric Scott Boltz) writes:

> As far as Intel CPUs go, aren't they the ones that expect multiple-byte
> data to have bytes swaped? - Yuck.

As far as I'm aware, nobody has *ever* made a machine that got byte-ordering
right in combination with several other crucial factors. This has had an
unpleasant effect on sofware quality.

Consistency is very important. The best starting point when designing a CPU
is this picture:

   +-------------------------------------------------+
   |    byte 1              |   byte 0               |
   +-------------------------------------------------+
     15 14 13 12 11 10 09 08  07 06 05 04 03 02 01 00
    2  2  2  2  2  2  2  2   2  2  2  2  2  2  2  2

where the numbers below denote the numerical value of the bits. Thus, the bit
ordering and the byte ordering are consistent with each other and with
standard mathematics. In this regard, the Intel was *right*, however goofy
they were about everything else. With this method, you can draw consistent
machine-data diagrams and never have to worry about which set of rules you
need to be using. One of the crimes of the upper end of the 68000 family is
its flipping of bit ordering in the more advanced bit operations.

Now, extending the previous picture to strings,

   abcdefg...                ...mnopqrstuvwxyz
   more significant <--   --> less significant
   higher address   <--   --> lower address

which is *consistent*. Once again, no more worries about which set of rules
you are using now. In this way, Intel got it wrong, since they want you to
have the first byte of a string at the lowest address, but then, so does
Motorola, except their picture is flipped. This is *wrong*. It also wastes
some processor capability: string operations usually proceed from left to
right, but with either the Intel or Motorola picture, you can't use the
index of a decrement-branch instruction as an index into the string itself.

The design I suggest is very different from what we're used to, but it works
just as well in the machine, and better in our heads when we're writing
programs. That translates into fewer bugs.

0) All memory diagrams are written sideways,
   with the most-significant bit, byte, word, whatever, on the left.

1) The base address of a string, struct, array is its *end*, not its beginning.
   These are written in memory diagrams from left-to-right, as in (0).
   String indexing is consistent with the value of decrement-branch indices.

2) Post-decrement and pre-increment addressing modes are implemented,
   instead of post-increment and pre-decrement.
   Post-decrement is just as fast as post-increment, if you were worried.

3) Processor stacks grow from low memory addresses to high,
   to allow for easier string concatenations opposite the stack.

Yes, I know, it's weird, but only because *everybody* got it wrong,
and that's what we've become accustomed to. Yuck.

Steve abell@netcom.com

In article <1992Nov2.154050.23454@jhunix.hcf.jhu.edu> eboltz@jhunix.hcf.jhu.edu (Eric Scott Boltz) writes:
  ... valid stuff deleted ...
>I'm actually quite worried about Nov. 10th. One of the machines Sun will
>release is a Tsunami (sp?) based SPARC which delivers 28 SpecMarks for
>less thatn $6k. Even though OpenWindows SUCKS, it's hard to justify
>NeXT's 16 Specmark, no bus system for $8k.
>
>The new HP will offer even better MIPS/$.
>
>My impression of NeXT is that they aren't even close to releasing any
>type of RISC machine. I believe that the vast majority of their efforts
>have gone into NS486. It seems that NeXT is targeting the PC market
>rather than take on the Workstation market. Granted, the line between
>is rather blurred, but if NeXT is going to compete with Sun, SGI and HP 
>they better get a RISC box fast.
  ... more stuff deleted ...
>Eric

A note blew by here about the R4400, which clocks at 75MHz (150MHz internal),
has a 32KB on chip cache, and is pin for pin compatable with the R4000PC/SC/MC
chips, leaving a clear upgrade path if NeXT were to ship an R4000 system today.
A dual R4000MC for under $10K would be quite a design feat, but it would get
them on the RISC map with a clear, glowing future.  Speed claims are based on
simulation, and come out at SPECint89 of 95 and SPECfp89 of 126 for a single
processor R4400 system.  There is a short blurb about "design enhancements for
graphics applications" in the announcement as well.  The R4400s 113 SPECmark89
overall for a single CPU is quite respectable compared to Suns claim of
109 SPECmark89 for their dual processor SPARCstation 10, which had a (probably)
outdated preannouncement price of $39,995 (2 CPUs, 64MB, 8-bit color, 1GB disk).

NeXTstep on a 200 SPECmark dual CPU box for $10K.  Its the stuff rumors (and
sometimes dreams) are made of.  Can Jobs and his crew be THAT good?  -Steve
-- 
~~~~~~~~~~~~~~~~~~~~
Stephen E. Halpin                    halpin@vsg1.dsg.dec.com

"You might just have to waste your life just to live."  - Soul Asylum

In article <1992Nov2.170741.2092@jhunix.hcf.jhu.edu> eboltz@jhunix.hcf.jhu.edu (Eric Scott Boltz) writes:
>This week's Info World, "Reports from the Field" claims that NeXT has
>terminated the dual 88110 NRW and will go with the P5 (Pentium) from
>Intel in future machines.
>
>This sucks for a few reasons:
>1. We won't have a faster NeXT till the P5 ships.
>2. We're stuck with Intel.
>3. The NRW will be AS FAST AS A HIGH-END CLONE.

This last item is not neccesarily true.  Let's suppose NeXT goes with
the P5.  Remember that the NeXT OS, Mach, already has support for
multiprocessor systems.  Doesn't the P5 have enhanced support for
working in multiprocessor systems?  Can you imagine a four processor
P5 system, with one processor devoted to a realtime Display PostScript
environment, another devoted to a realtime sound server (including a
port pin-compatible with the current DSP port), and two more for
general purpose CPU tasks?  NeXT could put out a wide variety of
systems simply by changing the number of processors.
-- 
Doug DeJulio
ddj+@cmu.edu

In article <1992Nov2.170741.2092@jhunix.hcf.jhu.edu>, eboltz@jhunix.hcf.jhu.edu (Eric Scott Boltz) writes:
> This week's Info World, "Reports from the Field" claims that NeXT has
> terminated the dual 88110 NRW and will go with the P5 (Pentium) from
> Intel in future machines.
** I feel like crying!  I can't believe NeXT isn't jumping on the RISC
   bandwagon.  I hope this ain't true.  If it is, I hope NeXT makes
   the best of it.
** Can some knowledgable person at NeXT please comment?
** Well at least a NeXT Intel machine won't cost as much as a
   NeXT RISC would.

The reason why I'm making such a big fuss about this is that I've been
monitoring the NeXT netnews groups and I even placed a subscription to
NeXTWorld (I don't even own a NeXT yet!) just to see what NeXT's plan
for a RISC machine would be. I was definitely going to purchase one
when it would be available for audio specific processing.


> 
> This sucks for a few reasons:
Sooooooooooooooooo TRUE!!!!

> 1. We won't have a faster NeXT till the P5 ships.
> 2. We're stuck with Intel.
> 3. The NRW will be AS FAST AS A HIGH-END CLONE.
> 
> Basically, if you stick with NeXT you commit to NeXTSTEP 486.
> I was going to order a NeXTstaiton Turbo Color, but now it seems
> that a BUC (Butt-Ugly-Clone) is the best hardware investment for
> NeXTSTEP.
> 
> Hey, IRIX ain't so bad...maybe I should go for an R4000 machine after all.
> 
> Eric

In article <1992Nov2.170741.2092@jhunix.hcf.jhu.edu> eboltz@jhunix.hcf.jhu.edu (Eric Scott Boltz) writes:
>This week's Info World, "Reports from the Field" claims that NeXT has
>terminated the dual 88110 NRW and will go with the P5 (Pentium) from
>Intel in future machines.

This is great news, assuming it to be true. Intel is winning the mass desktop
computer market, and Motorola has very little hope of getting it back. On the
one hand you have RISC workstations which are way ahead of the 88110 and on 
the other hand you've got 486+'s selling well. The only people staying with
Motorola in the micro computer world seem to be Atari, Commodore and Apple (and
even they have this possible IBM/Apple Power PC type platform).
>
>This sucks for a few reasons:
>1. We won't have a faster NeXT till the P5 ships.
>2. We're stuck with Intel.
>3. The NRW will be AS FAST AS A HIGH-END CLONE.

1 and 2 are neither here nor there. 3 may be true now, it won't be in a year.
>
>Basically, if you stick with NeXT you commit to NeXTSTEP 486.
>I was going to order a NeXTstaiton Turbo Color, but now it seems
>that a BUC (Butt-Ugly-Clone) is the best hardware investment for
>NeXTSTEP.

Some compatibles are rather nice. Maybe not as nice as a real NeXT, but then
again I assume you'll be able to buy a real NeXT in the future. Who cares
what the processor is? I'd rather see one that has mass market penetration
and is accessible to students.
 
>
>Hey, IRIX ain't so bad...maybe I should go for an R4000 machine after all.

Ha...
 
Philip McDunnough
University of Toronto
philip@utstat.toronto.edu

fernandeza@merrimack.edu writes:

>** Well at least a NeXT Intel machine won't cost as much as a
>   NeXT RISC would.

  Why do you think this??  Maybe if you're talking about competition
between Intel clone boxes I'd believe you, but RISC chips like the
Tsunami, PowerPC etc will sell in the range of $150->$200.  The Pentium
will be at least $500 for the year after it becomes available.

-- 
------------------------------------------------------------------------------
Evan Torrie.  Stanford University, Class of 199?       torrie@cs.stanford.edu 
Embrace rationalism, reject superstition.           Break away from the past.
"...there is only one thing worse than having marketing people on a project,
 and that is having no marketing people on a project."  - Larry Tesler.

In article <1992Nov2.170741.2092@jhunix.hcf.jhu.edu>
	eboltz@jhunix.hcf.jhu.edu (Eric Scott Boltz) writes:
>This week's Info World, "Reports from the Field" claims that NeXT has
>terminated the dual 88110 NRW and will go with the P5 (Pentium) from
>Intel in future machines.

But, according to UNIX World, NeXTSTEP will be available from DG
for its AViiON pizza boxes.

"I told you so"

					-=EPS=-

>>>>> "Eric" == Eric Scott Boltz (eboltz@jhunix.hcf.jhu.edu)

Eric> 3. The NRW will be AS FAST AS A HIGH-END CLONE.

Not true. Go take a look at the Road-runner (aka Sun 386i). Compare it
to a 386 PClown. The Road-runner is *so* much faster, shows real
smooth performance. It was a nice box, actually faster than the Sun-3
line at the time. The SPARC killed it, but hey, they showed what you
can do with an Intel chip if you have the guts and skills. We've even
had Intel folks say that the Road-runner is the only machine ever
designed to really use the 386.

Intel chip =!= PClown.

/Lars
--
Lars Fischer, fischer@iesd.auc.dk | It takes an uncommon mind to think of
CS Dept., Aalborg Univ., DENMARK. | these things.  -- Calvin

In article <1992Nov3.035949.23338@nntpd.lkg.dec.com> halpin@vsg1.dsg.dec.com (Stephen E. Halpin) writes:
>In article <1992Nov2.154050.23454@jhunix.hcf.jhu.edu> eboltz@jhunix.hcf.jhu.edu (Eric Scott Boltz) writes:
>  ... valid stuff deleted ...
>
>A note blew by here about the R4400, which clocks at 75MHz (150MHz internal),
>has a 32KB on chip cache, and is pin for pin compatable with the R4000PC/SC/MC
>chips, leaving a clear upgrade path if NeXT were to ship an R4000 system today.
>A dual R4000MC for under $10K would be quite a design feat, but it would get
>them on the RISC map with a clear, glowing future.  Speed claims are based on
>simulation, and come out at SPECint89 of 95 and SPECfp89 of 126 for a single
>processor R4400 system.  There is a short blurb about "design enhancements for
>graphics applications" in the announcement as well.  The R4400s 113 SPECmark89
>overall for a single CPU is quite respectable compared to Suns claim of
>109 SPECmark89 for their dual processor SPARCstation 10, which had a (probably)
>outdated preannouncement price of $39,995 (2 CPU, 64MB, 8-bit color, 1GB disk).
>NeXTstep on a 200 SPECmark dual CPU box for $10K.  Its the stuff rumors (and
>sometimes dreams) are made of.  Can Jobs and his crew be THAT good?  -Steve

  Dream on.  For my money, there is no way Jobs and crew are going to
  go with a processor from a once-almost-bankrupt now-acquired company 
  which is a subsidiary of an arch-competitor (SGI).  The risc is way to
  high :-) (sorry for the BAD pun) 

  Jon Rosen

In article <1992Nov5.061437.24266@RedBrick.COM>, jfr@RedBrick.COM () writes:
|>   Dream on.  For my money, there is no way Jobs and crew are going to
|>   go with a processor from a once-almost-bankrupt now-acquired company 
|>   which is a subsidiary of an arch-competitor (SGI).  The risc is way to
|>   high :-) (sorry for the BAD pun) 

I read recently that someone has produced a R4400 clone, so going with the 
R4400 would be as good as going with sparc (except of course that the R4400
is a better chip). Multi-vendor, RISC, etc.
Anyone know anything more?

|>   Jon Rosen

	james