Re: [OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Martin Smith]

On 16/03/2021 12:20, songbird wrote:

Nicholas Geovanis wrote:

On Sun, Mar 14, 2021, 1:50 PM Stefan Monnier 
wrote:

...

FWIW And MIPS was there even a bit earlier with their R4000 (tho the
software support for it only appeared some years later: they first
wanted to have an installed base to which to deploy the software), which
I believe was the first 64bit microprocessor.


And the demise of the DEC Alpha was quite unfortunate. It was super-fast
and OSF/1 was rock-solid. But DEC lost the competitive bid on that project
and Sequent/Dynix, based on hundreds of 486 CPUs, won it. Now owned by IBM
and deep-sixed: They really bought the customer base instead.


   i wondered what happened to them, but didn't look into it.
when the university got rid of the mainframe we switched to
Sequent machines.  the two cabinets replaced the entire
floor of Univac hardware (and all the AC and power costs).
the other nice thing was not listening to those printers
hammering away.


when I was working in the Mullard stores in the 60's they had an 
enormous computer in a very large air conditioned hall about a mile from 
the factory, I dont know what it precisely was but it ran off punched 
tape, and in a side room at the stores we have what was called a line 
printer that printed out invoice/advice note pairs it really was like a 
machine gun printing a line at a time





--
Martin

Re: [OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[songbird]

Nicholas Geovanis wrote:
> On Sun, Mar 14, 2021, 1:50 PM Stefan Monnier 
> wrote:
...
>> FWIW And MIPS was there even a bit earlier with their R4000 (tho the
>> software support for it only appeared some years later: they first
>> wanted to have an installed base to which to deploy the software), which
>> I believe was the first 64bit microprocessor.
>
> And the demise of the DEC Alpha was quite unfortunate. It was super-fast
> and OSF/1 was rock-solid. But DEC lost the competitive bid on that project
> and Sequent/Dynix, based on hundreds of 486 CPUs, won it. Now owned by IBM
> and deep-sixed: They really bought the customer base instead.

  i wondered what happened to them, but didn't look into it.
when the university got rid of the mainframe we switched to
Sequent machines.  the two cabinets replaced the entire 
floor of Univac hardware (and all the AC and power costs).
the other nice thing was not listening to those printers
hammering away.


  songbird

Re: [OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[tomas]

On Tue, Mar 16, 2021 at 12:08:51AM +0200, Andrei POPESCU wrote:
> On Lu, 15 mar 21, 11:19:55, Stefan Monnier wrote:
> > 
> > Last I heard Debian works on the M1 already :-), but its Emacs package
> > doesn't :-(
> 
> No surprise considering Emacs is itself a full OS :p
> 
> (sorry, could not resist)

  https://www.informatimago.com/linux/emacs-on-user-mode-linux.html

:-)

Cheers
 - t


signature.asc
Description: Digital signature

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Sven Hartge]

Christian Groessler  wrote:
> On 3/15/21 10:47 PM, Andrei POPESCU wrote:
>> On Lu, 15 mar 21, 20:24:56, Sven Hartge wrote:

>>> (I still vividly remember using memmaker and manual ordering the drivers
>>> in config.sys and autoexec.bat to shave another 2KB from the lower
>>> memory so the IPX driver would fit so Doom would run.)

>> For me it was Warcraft :)

>> And for some game (possibly also Warcraft) I had to pretend having a
>> sound card by listing the driver in config.sys, otherwise it wouldn't
>> even start.

> For me it was "Worms".

> And I was using QEMM and Quarterdeck Manifest to get maximal memory in 
> the lower 640k.

Ooooh, look at Mr Fancy here, cheating with 3rd party products to get
ahead :)

I'll throw in the special "maximise XMS memory boot disk" I had for
Comanche because that game just *hated* emm386.exe but without EMM,
stuff like "LOADHIGH" to push drivers into the UMB was not available and
I was too lazy to add another branch to my already convoluted config.sys
boot menu.

S°

-- 
Sigmentation fault. Core dumped.

Re: [TOTALLY OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Dan Ritter]

Andrei POPESCU wrote: 
> On Lu, 15 mar 21, 17:21:39, Dan Ritter wrote:
> > 
> > At last report: normal desktop Ryzens (nothing with a G suffix
> > unless it also has a PRO marking) 
> 
> Do you have a reliable source for the lack of ECC support in G suffix 
> processors?
> 
> And why would it work for PRO processors instead?
> 
> I think it's unlikely AMD has 2 different cores for PRO and non-PRO, 
> it's more likely it either works for both or neither.


https://www.asrock.com/mb/AMD/X570%20Taichi/index.asp#Specification

I'm going to omit a bunch of details:

AMD Ryzen series CPUs (Vermeer) support ...  ECC & non-ECC, un-buffered memory*
- AMD Ryzen series CPUs (Matisse) support ... ECC & non-ECC, un-buffered memory*
- AMD Ryzen series APUs (Renoir) support ... ECC & non-ECC, un-buffered memory*
- AMD Ryzen series CPUs (Pinnacle Ridge) support ... ECC & non-ECC, un-buffered 
memory*
- AMD Ryzen series CPUs (Picasso) support non-ECC, un-buffered memory*

* For Ryzen Series CPUs (Picasso), ECC is only supported with * PRO CPUs.


The first APUs are the Raven Ridge, 2200G and 2400G, which
aren't even supported on the current motherboards

The next are Picassos, 3200G and 3400G, there's an explicit
statement that only the PRO versions support ECC.

The current ones are Renoir, 4000 series, and I haven't got a
reliable source that they are ECC only on the PRO -- but I
strongly suspect it.

It's not the cores that differ between the PROs and non- -- it's
the I/O chiplet.

-dsr-

Re: [OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Roberto C . Sánchez]

On Tue, Mar 16, 2021 at 12:08:51AM +0200, Andrei POPESCU wrote:
> On Lu, 15 mar 21, 11:19:55, Stefan Monnier wrote:
> > 
> > Last I heard Debian works on the M1 already :-), but its Emacs package
> > doesn't :-(
> 
> No surprise considering Emacs is itself a full OS :p
> 
Yeah, but it could really do with a decent text editor :-p

> (sorry, could not resist)
> 
(neither could I)

Regards,

-Roberto

-- 
Roberto C. Sánchez

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Christian Groessler]

On 3/15/21 10:47 PM, Andrei POPESCU wrote:

On Lu, 15 mar 21, 20:24:56, Sven Hartge wrote:

(I still vividly remember using memmaker and manual ordering the drivers
in config.sys and autoexec.bat to shave another 2KB from the lower
memory so the IPX driver would fit so Doom would run.)

For me it was Warcraft :)

And for some game (possibly also Warcraft) I had to pretend having a
sound card by listing the driver in config.sys, otherwise it wouldn't
even start.



For me it was "Worms".

And I was using QEMM and Quarterdeck Manifest to get maximal memory in 
the lower 640k.


regards,
chris

Re: [OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Andrei POPESCU]

On Lu, 15 mar 21, 11:19:55, Stefan Monnier wrote:
> 
> Last I heard Debian works on the M1 already :-), but its Emacs package
> doesn't :-(

No surprise considering Emacs is itself a full OS :p

(sorry, could not resist)

Kind regards,
Andrei
-- 
http://wiki.debian.org/FAQsFromDebianUser


signature.asc
Description: PGP signature

Re: [TOTALLY OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Andrei POPESCU]

On Lu, 15 mar 21, 17:21:39, Dan Ritter wrote:
> 
> At last report: normal desktop Ryzens (nothing with a G suffix
> unless it also has a PRO marking) 

Do you have a reliable source for the lack of ECC support in G suffix 
processors?

And why would it work for PRO processors instead?

I think it's unlikely AMD has 2 different cores for PRO and non-PRO, 
it's more likely it either works for both or neither.

Kind regards,
Andrei
-- 
http://wiki.debian.org/FAQsFromDebianUser


signature.asc
Description: PGP signature

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Andrei POPESCU]

On Lu, 15 mar 21, 20:24:56, Sven Hartge wrote:
> 
> (I still vividly remember using memmaker and manual ordering the drivers
> in config.sys and autoexec.bat to shave another 2KB from the lower
> memory so the IPX driver would fit so Doom would run.)

For me it was Warcraft :)

And for some game (possibly also Warcraft) I had to pretend having a 
sound card by listing the driver in config.sys, otherwise it wouldn't 
even start.

Kind regards,
Andrei
-- 
http://wiki.debian.org/FAQsFromDebianUser


signature.asc
Description: PGP signature

Re: [TOTALLY OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Dan Ritter]

Anssi Saari wrote: 
> Dan Ritter  writes:
> 
> As for the ECC support in Ryzen CPUs, as I understand it it's a bit of a
> mess. Sure the CPUs support it but if it's not validated by motherboard
> manufacturers, how do you know it actually works reliably?

... by trying it out and reporting the results to others, and
reading their results and reporting your confirmation.

This isn't a thing that the motherboard manufacturer can put in
by accident.

Anyway. If you need ECC support, you buy an EPYC server and get
registered ECC support. If you would like to have ECC as a feature, you
get a Ryzen board that's reported to work, and you get
unbuffered ECC for one-bit correction and two-bit reporting.

Then you overclock it to generate RAM errors, and it shows up in
your system log. Then you bring it back down to normal speed.

At last report: normal desktop Ryzens (nothing with a G suffix
unless it also has a PRO marking) on any ASrock, most ASUS, and
some Gigabyte motherboards will support this. To the best of my
current knowledge, no MSI motherboards.

-dsr-

Re: [TOTALLY OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Anssi Saari]

Dan Ritter  writes:

> Intel knew that their argument was bull: they owned the market
> and needed ways of subdividing their CPUs to fit every price
> point. Turning off ECC support was one of those ways.

> That strategy started with the 80486, when they brought out a
> cheap version called the 80486SX which lacked a floating point
> unit. The SX has the floating point unit, it was just turned
> off.

Initially, yes. A panic move when AMD brought out their 40 MHz 386. It
worked, got popular and later on the 486SX was manufactured separately
with a smaller die and no floating point.

As for the ECC support in Ryzen CPUs, as I understand it it's a bit of a
mess. Sure the CPUs support it but if it's not validated by motherboard
manufacturers, how do you know it actually works reliably?

Re: [TOTALLY OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Dan Ritter]

Sven Hartge wrote: 
> Stefan Monnier  wrote:
> 
> > From a purely technical perspective, it's hard to understand how Intel
> > managed to pour so much energy into such an obviously bad idea.  The
> > only explanations seem all to be linked to market strategies.
> 
> This history repeats for Intel on several fronts:
> 
> Or the discussion about ECC for desktop devices. Intel argues "not
> needed", which is, if you follow the Rowhammer issues, not true. AMD
> just does it and it works.

Intel knew that their argument was bull: they owned the market
and needed ways of subdividing their CPUs to fit every price
point. Turning off ECC support was one of those ways.

That strategy started with the 80486, when they brought out a
cheap version called the 80486SX which lacked a floating point
unit. The SX has the floating point unit, it was just turned
off. Worse: purchasing the 80487 math coprocessor to enable
floating point support... the 487 was a full 486, that turned
off the original.

> Then there was FB-DIMM back in the 2008s. Nice idea, just, again, too
> expensive and disconnected from the market in the end.

Intel wanted more pricing points. 

> I personally am really glad that AMD got their stuff together again and
> with their ZenX-Architectures showed Intel how it is done.
> 
> What AMD now needs is a hit in the low, lower and ultra-low power
> segment.

They've got the low and lower parts now: 35W and 15W 4000-series
APUs, from the Renoir design. Stefan and I were just talking
about how you can't buy one with a normal motherboard right now
because they are entirely allocated to systems integrators. AMD
is selling 100% of production.

They don't have any 7W or lower parts, but those things aren't
very interesting compared to ARM64 architecture, where Qualcomm
and Apple and any number of smaller shops are doing great things
in the tablet and phone space.

-dsr-

Re: [TOTALLY OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Sven Hartge]

Stefan Monnier  wrote:

> From a purely technical perspective, it's hard to understand how Intel
> managed to pour so much energy into such an obviously bad idea.  The
> only explanations seem all to be linked to market strategies.

This history repeats for Intel on several fronts:

Look at the Netburst Pentium 4 desaster, which as scrapped as soon as
the Israel division showed their improved concept based on the P3, which
ran laps around the P4 while at the same time using far less power and
had a bigger yield.

Or the discussion about ECC for desktop devices. Intel argues "not
needed", which is, if you follow the Rowhammer issues, not true. AMD
just does it and it works.

Then there was FB-DIMM back in the 2008s. Nice idea, just, again, too
expensive and disconnected from the market in the end.

And all in all the rather slow improvments on the CPU fronts, the
piecemeal 5% increases sold as "big achievements" every year, while at
the same time all improvements turned out to be major security problems.

I personally am really glad that AMD got their stuff together again and
with their ZenX-Architectures showed Intel how it is done.

What AMD now needs is a hit in the low, lower and ultra-low power
segment.

Grüße,
S°

-- 
Sigmentation fault. Core dumped.

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Sven Hartge]

Joe  wrote:
> On Mon, 15 Mar 2021 12:34:42 +0100 Sven Hartge  wrote:

>> Imagine a PC with 4GB adressable memory space in 1980.

> I can. It would have cost as much as a mainframe to make full use of it.

I don't say to put it in, only to have a flat 32bit address range.

Just like the current 64bit systems don't have 16 Exabyte of memory in
them.

(I still vividly remember using memmaker and manual ordering the drivers
in config.sys and autoexec.bat to shave another 2KB from the lower
memory so the IPX driver would fit so Doom would run.)

S°

-- 
Sigmentation fault. Core dumped.

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Stefan Monnier]

>> No it wouldn't, and we had it by the late '80's with the advent of
>> 68040 abd 68060 accellerator boards for the Amiga's. But that flat
>> memory model and poor production QC doomed it.  Any program could make
>> a missfire and write into another programs memory space, crashing the
>> whole Mary Ann.
> Starting in '82 the 68010 added virtual memory and virtualization suport.

[ I can't remember any discussion of virtualization for that.
  Back then this only existed on things like IBM mainframes and noone in
  the workstation-and-lower markets cared about it, AFAIK.  ]

Note that this is only true in the sense of "wifi ready" (a laptop that
came without any wifi card but maybe with an antenna in the bezel): the
68010 was a very minor improvement of the 68000 which just fixed some
blunders that were making it (almost) impossible to provide support for
virtual memory.  You needed additional help (like an MMU) in order to
get virtual memory on the 68010 and that usually ended up very costly in
terms of performance.

Virtual memory only became vaguely usable with the 68020 (and then
actually usable on the 68030).


Stefan

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Gene Heskett]

On Monday 15 March 2021 12:40:51 John Hasler wrote:

> Gene writes:
> > No it wouldn't, and we had it by the late '80's with the advent of
> > 68040 abd 68060 accellerator boards for the Amiga's. But that flat
> > memory model and poor production QC doomed it.  Any program could
> > make a missfire and write into another programs memory space,
> > crashing the whole Mary Ann.
>
> Starting in '82 the 68010 added virtual memory and virtualization
> suport.

But by then the amiga design was frozen until the funeral.

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: [OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Michael Stone]

On Mon, Mar 15, 2021 at 11:03:59AM -0400, Stefan Monnier wrote:

From a purely technical perspective, it's hard to understand how Intel
managed to pour so much energy into such an obviously bad idea.
The only explanations seem all to be linked to market strategies.


They just had too much easy money coming in from the windows/x86 desktop 
monopoly. It took years before they really had to justify in a critical 
way the money they were spending. 

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[John Hasler]

Gene writes:
> No it wouldn't, and we had it by the late '80's with the advent of
> 68040 abd 68060 accellerator boards for the Amiga's. But that flat
> memory model and poor production QC doomed it.  Any program could make
> a missfire and write into another programs memory space, crashing the
> whole Mary Ann.

Starting in '82 the 68010 added virtual memory and virtualization suport.
-- 
John Hasler 
jhas...@newsguy.com
Elmwood, WI USA

Re: [OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Dan Ritter]

Stefan Monnier wrote: 
> > There's already work in progress to port Linux mainline (and 
> > consequently Debian) to the Apple M1 :)
> 
> Since the M1 implements the ARM instruction set, I don't think there's
> much work to do here, indeed (most likely the hardest part is to fight
> Apple's opaqueness).
> 
> Last I heard Debian works on the M1 already :-), but its Emacs package
> doesn't :-(

Graphics is currently the blocker. Framebuffer works, but
getting the GPU working beyond that will probably be fun for
someone.

https://asahilinux.org/2021/03/progress-report-january-february-2021/

contains lots of useful info.

-dsr-

Re: [OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Nicholas Geovanis]

On Sun, Mar 14, 2021, 1:50 PM Stefan Monnier 
wrote:

> > Well, nearly. Itanium Merced was 2001 [1] (althoug you wouldn't buy
> > /that/ as a private person), DEC Alpha was even 1992 [2];
>
> FWIW And MIPS was there even a bit earlier with their R4000 (tho the
> software support for it only appeared some years later: they first
> wanted to have an installed base to which to deploy the software), which
> I believe was the first 64bit microprocessor.
>

And the demise of the DEC Alpha was quite unfortunate. It was super-fast
and OSF/1 was rock-solid. But DEC lost the competitive bid on that project
and Sequent/Dynix, based on hundreds of 486 CPUs, won it. Now owned by IBM
and deep-sixed: They really bought the customer base instead.

The final pedantry is that, contrary to an earlier post, the first IBM PCs
were built around the 8088, not the 8086.

IIRC the claim back then was that adding 64bit support to the R4000 was
> rather cheap (it increased the die area by a few percents only, and
> 64bit adds were still fast enough not to slow down the overall chip's
> frequency).
>
> The same must have been true for the Opterons (except that the increase
> in die area much have been even much smaller since the CPU itself had
> become a much smaller part of the overall die because of the
> incorporation of things like the memory controller and the L1 and L2
> caches).
>
> So it was a great move on the part of AMD: cheap to implement but with
> an enormous marketing impact.
>
>
> Stefan
>
>

Re: [OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Stefan Monnier]

>> The original plan/claims was that the support for legacy i386
>> application would be "just as fast".  This never materialized
>> (unsurprisingly: it's easy to make a CPU that can run efficiency several
>> slightly different instruction sets (ISA), like your average amd64 CPU which
>> can run applications using the amd64 ISA, the i386 ISA, the 80286 ISA
>> or the 8086 ISA, more or less; but it's much harder to make a CPU that
>> can run efficiently very different ISAs).
> Apple seems to be doing quite well with the M1.

But that's not a CPU that runs amd64 code: the amd64 code is executed on
it by software emulation rather than by hardware emulation.  And indeed,
Intel could have developed an efficient software emulation of amd64 for
its Itanium which could have been more efficient than its own
hardware emulator.

[ Similarly, at some point in time, DEC's Alpha was claimed to be the
  fastest processor to run i386 code, via its software emulation. ;-)  ]

Apple has a lot of experience in that kind of emulation (having done it
for the transition from Motorola's 68K to PowerPC, then again from
PowerPC to i386, and now from amd64 to ARM (notice they relied on
hardware emulation for the i386 to amd64 transition)).

But note that they only do emulation for applications AFAIK, which is
easier than doing a "full" emulation that lets you run an actual OS
(like `qemu` does).

> There's already work in progress to port Linux mainline (and 
> consequently Debian) to the Apple M1 :)

Since the M1 implements the ARM instruction set, I don't think there's
much work to do here, indeed (most likely the hardest part is to fight
Apple's opaqueness).

Last I heard Debian works on the M1 already :-), but its Emacs package
doesn't :-(


Stefan

Re: [OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Stefan Monnier]

>> Indeed.  Also, they wanted to move away from the i386 instruction set
>> so as not to be bothered by pre-existing licensing agreements with
>> AMD, and thus making sure there'd be no competing implementation.  The
>> IA64 architecture was quite complex, and there are reasons to believe
>> that complexity was seen as a virtue (makes it easier to get more
>> patents and keep competitors out).
> HP then also poured additional stuff into the architecture to make
> migration from PA-RISC easier.  I imagine this also made stuff vastly
> more complex.

It has all the signs of a "design by committee" were you get the union
of all the ideas, indeed :-(

But I think for such a thing to get the time and funding needed to get
to production, there needs to be a commitment to the idea that such
complexity is good.

> I think, IBM is big enough and old enough and established enough with
> POWER that a "young whippersnapper" like Intel is no real danger to them
> in their own enclosed Mainframe walled garden. I believe Apple moving
> away from PowerPC did more damage to IBMs aspirations in that market.

Agreed.

> For the others: they where either on board from the start (like HP),
> where already dead (like DEC/Compaq) or slipping into the embedded
> market (like MIPS).

I didn't want to imply that they would have survived (that slice of the
CPU market was shrinking fast anyway: after the Pentium Pro, they were
not noticeably faster than PCs any more and the market was too small to
keep financing the development of leading CPUs, especially since for
high-end machines all the value was in the interconnect rather than the
CPUs anyway), but the IA64 was explicitly the end of it for them (and
that happened long before the first IA64 CPU was available).

> And SPARC: after being bought by Oracle, the end was more or less
> directly clear.

But that took place much later: the IA64 buzz that killed Alpha/PA/MIPS
was in the 90s whereas Oracle bought SPARC in 2009.

> Indeed. The German computer magazine c't had many interesting articles
> about the IA64 architecture and also quite early painted its dark
> future, because of ever slipping sales figures, performance problems,
> the failure to deliver on made promises and the increasing pressure of
> the i386/amd64 architectures.

>From a purely technical perspective, it's hard to understand how Intel
managed to pour so much energy into such an obviously bad idea.
The only explanations seem all to be linked to market strategies.


Stefan

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Michael Stone]

On Mon, Mar 15, 2021 at 12:53:46PM +, Joe wrote:

On Mon, 15 Mar 2021 12:34:42 +0100 Sven Hartge  wrote:

Imagine a PC with 4GB adressable memory space in 1980.


I can. It would have cost as much as a mainframe to make full use of it.


More. Memory was often the largest line item back then, and ordinary 
mainframes couldn't afford much of it. The Cray 2 was a game-changer in 
the supercomputer space with its 1Gbyte memory capacity. Mostly those 
were bought by three letter agencies, but some really large corporations 
and universities with very generous donors got one.

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[tomas]

On Mon, Mar 15, 2021 at 10:45:15AM -0400, Stefan Monnier wrote:
> >> Another rumor I read was that IBM, when developing the first IBM PC in
> >> 1980, opted to use the 8086/8088 CPU instead of the also availble M68k
> >> CPU because the Intel one was less powerful so it would not be in
> >> competition with the mainframes the PC was supposed to interface with
> >> primarily.
> > Too lazy to research now, but it sounds credible, yes.
> 
> I'm sure there have been several different factors and it's hard to know
> which were more important (often the more personal and less technical
> factors are the more important ones in those areas, but the hardest to
> track down and verify).

ISTR that the Big Iron and the small stuff factions whithin IBM were
in fierce competition at the time. That's why the idea seemed plausible
to me.

[...]  Another important factor (linked to pragmatic
> constraints of overall production cost and availability of all the
> various components at particular dates) made it important to use an 8bit
> interface between the CPU and the system (which arguably also ensured it
> was no threat performancewise to the rest of IBM's linup).
> That's another reason why they went with the 8088 rather than the 8086,
> and also another reason why they went with Intel rather than Motorola,
> since the 68008 wasn't available yet.

...the outcome was surely that of multiple factors. IBM was a complex
beast at the time!

Cheers
 - t


signature.asc
Description: Digital signature

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Stefan Monnier]

>> Another rumor I read was that IBM, when developing the first IBM PC in
>> 1980, opted to use the 8086/8088 CPU instead of the also availble M68k
>> CPU because the Intel one was less powerful so it would not be in
>> competition with the mainframes the PC was supposed to interface with
>> primarily.
> Too lazy to research now, but it sounds credible, yes.

I'm sure there have been several different factors and it's hard to know
which were more important (often the more personal and less technical
factors are the more important ones in those areas, but the hardest to
track down and verify).  Another important factor (linked to pragmatic
constraints of overall production cost and availability of all the
various components at particular dates) made it important to use an 8bit
interface between the CPU and the system (which arguably also ensured it
was no threat performancewise to the rest of IBM's linup).
That's another reason why they went with the 8088 rather than the 8086,
and also another reason why they went with Intel rather than Motorola,
since the 68008 wasn't available yet.


Stefan

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[John Hasler]

Gene writes:
> That, IIRC was a new, super shiny, thing from zilog. No experience
> with it, but if it was as unreliable as the z-80, was, I'm not sorry
> it failed. The Z-80 had an instruction that swapped the
> foregrund/background register sets.  But it only worked on odd hours
> of the day. And had no way of testing if the command had worked
> without sacrificing 1 of the three registers. in both sets.

I used lots of Z80s and had good luck with them.  I wrote an OS for my
first Z80 homebrew computer that used register swapping to service
interrupts and print in the background.  It worked quite well.  Most
applications used only one register set, though, due to the need for
Intel compatibility.

My first Unix machine was an Onyx with a Z8000 running System III.  The
8 inch disk got flaky after about ten years but other than that it was
quite reliable.  Odd architecture, though.  I would have preferred 68k.
-- 
John Hasler 
jhas...@newsguy.com
Elmwood, WI USA

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[tomas]

On Mon, Mar 15, 2021 at 10:02:12AM -0400, Gene Heskett wrote:

[...]

> Snerk. We all did that back in the day, Tomas. that and similar magazines 
> were this 8th grade graduates electronics education. Do they still exist 
> today? Retired now, so the subs expired.

Some of them: https://www.ee.com/

Cheers
 - t


signature.asc
Description: Digital signature

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Gene Heskett]

On Monday 15 March 2021 09:53:40 to...@tuxteam.de wrote:

> On Mon, Mar 15, 2021 at 09:31:05AM -0400, Gene Heskett wrote:
> > On Monday 15 March 2021 07:05:02 to...@tuxteam.de wrote:
> > > On Mon, Mar 15, 2021 at 11:09:35AM +0100, Sven Hartge wrote:
> > >
> > > [...]
> > >
> > > > Another rumor I read was that IBM, when developing the first IBM
> > > > PC in 1980, opted to use the 8086/8088 CPU instead of the also
> > > > availble M68k CPU because the Intel one was less powerful so it
> > > > would not be in competition with the mainframes the PC was
> > > > supposed to interface with primarily.
> > >
> > > Too lazy to research now, but it sounds credible, yes.
> > >
> > > > If this rumor is true and IBM had acted differently, the PC
> > > > ecosystem today would also look quite differently.
> > >
> > > Or the Z8000. Absolutely. 8086 was, architecturally, the worst
> > > possible choice at that time.
> >
> > That, IIRC was a new, super shiny, thing from zilog. No experience
> > with it, but if it was as unreliable as the z-80, was, I'm not sorry
> > it failed. The Z-80 had an instruction that swapped the
>
> [...]
>
> I take that back. Z8000 was a 16 bit data/24 bit address thing; it
> did have a segmented architecture, so it wasn't as "clean" as I
> remembered it. At that time I was just a little student, so my
> "experience" with that stuff was to drool over design articles
> in the usual magazines (EE, AFAIR).
>
> Cheers
>  - t
Snerk. We all did that back in the day, Tomas. that and similar magazines 
were this 8th grade graduates electronics education. Do they still exist 
today? Retired now, so the subs expired.

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[tomas]

On Mon, Mar 15, 2021 at 09:31:05AM -0400, Gene Heskett wrote:
> On Monday 15 March 2021 07:05:02 to...@tuxteam.de wrote:
> 
> > On Mon, Mar 15, 2021 at 11:09:35AM +0100, Sven Hartge wrote:
> >
> > [...]
> >
> > > Another rumor I read was that IBM, when developing the first IBM PC
> > > in 1980, opted to use the 8086/8088 CPU instead of the also availble
> > > M68k CPU because the Intel one was less powerful so it would not be
> > > in competition with the mainframes the PC was supposed to interface
> > > with primarily.
> >
> > Too lazy to research now, but it sounds credible, yes.
> >
> > > If this rumor is true and IBM had acted differently, the PC
> > > ecosystem today would also look quite differently.
> >
> > Or the Z8000. Absolutely. 8086 was, architecturally, the worst
> > possible choice at that time.
> >
> That, IIRC was a new, super shiny, thing from zilog. No experience with 
> it, but if it was as unreliable as the z-80, was, I'm not sorry it 
> failed. The Z-80 had an instruction that swapped the 

[...]

I take that back. Z8000 was a 16 bit data/24 bit address thing; it
did have a segmented architecture, so it wasn't as "clean" as I
remembered it. At that time I was just a little student, so my
"experience" with that stuff was to drool over design articles
in the usual magazines (EE, AFAIR).

Cheers
 - t


signature.asc
Description: Digital signature

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Gene Heskett]

On Monday 15 March 2021 08:53:46 Joe wrote:

> On Mon, 15 Mar 2021 12:34:42 +0100
>
> Sven Hartge  wrote:
> > to...@tuxteam.de wrote:
> > > On Mon, Mar 15, 2021 at 11:09:35AM +0100, Sven Hartge wrote:
> > >> Another rumor I read was that IBM, when developing the first IBM
> > >> PC in 1980, opted to use the 8086/8088 CPU instead of the also
> > >> availble M68k CPU because the Intel one was less powerful so it
> > >> would not be in competition with the mainframes the PC was
> > >> supposed to interface with primarily.
> > >
> > > Too lazy to research now, but it sounds credible, yes.
> > >
> > >> If this rumor is true and IBM had acted differently, the PC
> > >> ecosystem today would also look quite differently.
> > >
> > > Or the Z8000. Absolutely. 8086 was, architecturally, the worst
> > > possible choice at that time.
> >
> > Having had a 68k would have been awesome. No stupid memory
> > segmentation, 32bit instructions and internal address size, 24bit
> > external address size.
> >
> > Imagine a PC with 4GB adressable memory space in 1980.
>
> I can. It would have cost as much as a mainframe to make full use of
> it.

No it wouldn't, and we had it by the late '80's with the advent of 68040 
abd 68060 accellerator boards for the Amiga's. But that flat memory  
model and poor production QC doomed it.  Any program could make a 
missfire and write into another programs memory space, crashing the 
whole Mary Ann. Then Commode-door brought out a 68060 board for the 
4000's. Major failure because that $1600, 4 square inches of pcb, had 
every electrolytic capacitor installed in reverse polarity. Too damned 
compact to be easily fixed, but I did two of them anyway.

Yup, I am a card carrying CET. What else could I do?

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Gene Heskett]

On Monday 15 March 2021 07:05:02 to...@tuxteam.de wrote:

> On Mon, Mar 15, 2021 at 11:09:35AM +0100, Sven Hartge wrote:
>
> [...]
>
> > Another rumor I read was that IBM, when developing the first IBM PC
> > in 1980, opted to use the 8086/8088 CPU instead of the also availble
> > M68k CPU because the Intel one was less powerful so it would not be
> > in competition with the mainframes the PC was supposed to interface
> > with primarily.
>
> Too lazy to research now, but it sounds credible, yes.
>
> > If this rumor is true and IBM had acted differently, the PC
> > ecosystem today would also look quite differently.
>
> Or the Z8000. Absolutely. 8086 was, architecturally, the worst
> possible choice at that time.
>
That, IIRC was a new, super shiny, thing from zilog. No experience with 
it, but if it was as unreliable as the z-80, was, I'm not sorry it 
failed. The Z-80 had an instruction that swapped the 
foregrund/background register sets.  But it only worked on odd hours of 
the day. And had no way of testing if the command had worked without 
sacrificing 1 of the three registers. in both sets.

When I finally got schmardter and wrote a test loop to check it, called 
zilog, and it was out of their 90 day warranty. They would not replace 
it. I should have called them and got a sample, but I'm honest and told 
them the truth. I never again used a zilog chip in anything.

I was then on a small town AM/FM radio stations budget, developing an 
Automatic Transmitter System for a temperature picky fm transmitter that 
really ought to have been replaced, starting with the brand label on the 
front panel.

This was in 1980, and the late 70's saw many Ma and Pa small town 
broadcasters severely impacted by trying to replace aging tube 
transmitters with early solid state versions before the tech was mature 
enough to be as dependable as the tube models. It took another ten years 
before the semi failure rates went below that of electrolytic 
capacitors.

Now design rule violations by the gear makers are responsible for a good 
share of the failure bugs. But they are a distant part of the list, well 
behind electrolytic caps whose technology has not been seriously 
improved in a hundred years now.  Even Tesla has put money into new 
versions, and come up short or they would be in his cars replaceing the 
lithium and dangerous batteries right now.

> Cheers
>  - t
Take care and stay safe and well, Tomas.

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Joe]

On Mon, 15 Mar 2021 12:34:42 +0100
Sven Hartge  wrote:

> to...@tuxteam.de wrote:
> > On Mon, Mar 15, 2021 at 11:09:35AM +0100, Sven Hartge wrote:  
> 
> >> Another rumor I read was that IBM, when developing the first IBM PC
> >> in 1980, opted to use the 8086/8088 CPU instead of the also
> >> availble M68k CPU because the Intel one was less powerful so it
> >> would not be in competition with the mainframes the PC was
> >> supposed to interface with primarily.  
> 
> > Too lazy to research now, but it sounds credible, yes.  
> 
> >> If this rumor is true and IBM had acted differently, the PC
> >> ecosystem today would also look quite differently.  
> 
> > Or the Z8000. Absolutely. 8086 was, architecturally, the worst
> > possible choice at that time.  
> 
> Having had a 68k would have been awesome. No stupid memory
> segmentation, 32bit instructions and internal address size, 24bit
> external address size.
> 
> Imagine a PC with 4GB adressable memory space in 1980.
> 
>
I can. It would have cost as much as a mainframe to make full use of it.

-- 
Joe

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[IL Ka]

>
>
> No stupid memory segmentation,
>

IMHO segmentation was a good idea originally.
You could have separate segments for code and data and since 286 it is
possible to protect them (AFAIK segments were also used to separate
user-space and kernel-space)
But with the advent of virtual memory (386), they have become an obsolete
legacy.

Intel is full of such things: hardware context switching, some old MMX
instructions, I/O ports, real mode: nobody needs all of that in 2021, but
it exists and occupies place in "intel development manual"

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[tomas]

On Mon, Mar 15, 2021 at 12:34:42PM +0100, Sven Hartge wrote:

[...]

> Having had a 68k would have been awesome. No stupid memory segmentation,

So were Z8000, NS32K and many others. The horrible segmentation thing on
the '86 were the tribute to backward compatibility, which is the price
you pay for market dominance :-)

> 32bit instructions and internal address size, 24bit external address size.
> 
> Imagine a PC with 4GB adressable memory space in 1980.

Yup.

Cheers
 - t


signature.asc
Description: Digital signature

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Sven Hartge]

to...@tuxteam.de wrote:
> On Mon, Mar 15, 2021 at 11:09:35AM +0100, Sven Hartge wrote:

>> Another rumor I read was that IBM, when developing the first IBM PC
>> in 1980, opted to use the 8086/8088 CPU instead of the also availble
>> M68k CPU because the Intel one was less powerful so it would not be
>> in competition with the mainframes the PC was supposed to interface
>> with primarily.

> Too lazy to research now, but it sounds credible, yes.

>> If this rumor is true and IBM had acted differently, the PC ecosystem
>> today would also look quite differently.

> Or the Z8000. Absolutely. 8086 was, architecturally, the worst possible
> choice at that time.

Having had a 68k would have been awesome. No stupid memory segmentation,
32bit instructions and internal address size, 24bit external address size.

Imagine a PC with 4GB adressable memory space in 1980.

S°

-- 
Sigmentation fault. Core dumped.

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[tomas]

On Mon, Mar 15, 2021 at 11:09:35AM +0100, Sven Hartge wrote:

[...]

> Another rumor I read was that IBM, when developing the first IBM PC in
> 1980, opted to use the 8086/8088 CPU instead of the also availble M68k
> CPU because the Intel one was less powerful so it would not be in
> competition with the mainframes the PC was supposed to interface with
> primarily.

Too lazy to research now, but it sounds credible, yes.

> If this rumor is true and IBM had acted differently, the PC ecosystem
> today would also look quite differently.

Or the Z8000. Absolutely. 8086 was, architecturally, the worst possible
choice at that time.

Cheers
 - t


signature.asc
Description: Digital signature

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Sven Hartge]

to...@tuxteam.de wrote:
> On Mon, Mar 15, 2021 at 09:15:10AM +0100, Sven Hartge wrote:

>> For the others: they where either on board from the start (like HP),
>> where already dead (like DEC/Compaq) or slipping into the embedded
>> market (like MIPS).

> MIPS had its chance to become the unified architecture for high-end
> workstations [1]. Until it was bought up by Silicon Graphics (SGI).
> Which, on the one hand was bitterly needed by MIPS, because they
> needed that cash injection, and by SGI, because they depended on the
> MIPS architecture.

> On the other hand, though, all other workstation developers, in fierce
> competition with SGI, didn't want /that/ dependency and went to look
> for/make other architectures (Power, Alpha, PA, you name it).

> So on the one hand, we might have, these days, been running on MIPS;
> on that other hand, we wouldn't have ARM, and -- who knows, soon,
> Risc-V. And Linus Torvalds wouldn't have had this cool stint at
> Transmeta. But that is a totally different kettle of fish. 

Another rumor I read was that IBM, when developing the first IBM PC in
1980, opted to use the 8086/8088 CPU instead of the also availble M68k
CPU because the Intel one was less powerful so it would not be in
competition with the mainframes the PC was supposed to interface with
primarily.

If this rumor is true and IBM had acted differently, the PC ecosystem
today would also look quite differently.

Grüße,
Sven.

-- 
Sigmentation fault. Core dumped.

Re: [EVEN MORE OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[tomas]

On Mon, Mar 15, 2021 at 09:15:10AM +0100, Sven Hartge wrote:

[...]

> For the others: they where either on board from the start (like HP),
> where already dead (like DEC/Compaq) or slipping into the embedded
> market (like MIPS).

MIPS had its chance to become the unified architecture for high-end
workstations [1]. Until it was bought up by Silicon Graphics (SGI).
Which, on the one hand was bitterly needed by MIPS, because they
needed that cash injection, and by SGI, because they depended on the
MIPS architecture.

On the other hand, though, all other workstation developers, in fierce
competition with SGI, didn't want /that/ dependency and went to look
for/make other architectures (Power, Alpha, PA, you name it).

So on the one hand, we might have, these days, been running on MIPS;
on that other hand, we wouldn't have ARM, and -- who knows, soon,
Risc-V. And Linus Torvalds wouldn't have had this cool stint at
Transmeta. But that is a totally different kettle of fish. 

Or is it?

> -- 
> Sigmentation fault. Core dumped.

:-)

Reminds me of an error message somewhere deep in TeX's or
METAFONT's bowels (sorry, from memory, therefore imprecise)
asking for "...someone to fix me fix me".

Cheers

[1] https://en.wikipedia.org/wiki/Advanced_Computing_Environment

 - t


signature.asc
Description: Digital signature

Re: [OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Andrei POPESCU]

On Du, 14 mar 21, 15:17:39, Stefan Monnier wrote:
> 
> The original plan/claims was that the support for legacy i386
> application would be "just as fast".  This never materialized
> (unsurprisingly: it's easy to make a CPU that can run efficiency several
> slightly different instruction sets (ISA), like your average amd64 CPU which
> can run applications using the amd64 ISA, the i386 ISA, the 80286 ISA
> or the 8086 ISA, more or less; but it's much harder to make a CPU that
> can run efficiently very different ISAs).

Apple seems to be doing quite well with the M1. Apparently it has a few 
custom instructions to speed up x86 emulation. They also have the 
benefit of controlling the software and now also the hardware stack.

There's already work in progress to port Linux mainline (and 
consequently Debian) to the Apple M1 :)

Kind regards,
Andrei
-- 
http://wiki.debian.org/FAQsFromDebianUser


signature.asc
Description: PGP signature

Re: [OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Sven Hartge]

Stefan Monnier  wrote:

>> Note: when IA64 was designed (starting in 1994 at HP) we where nowhere
>> near the limits of the 32bit i386 architecture with RAM and frequency,
>> so it made sense, somewhat.

> Indeed.  Also, they wanted to move away from the i386 instruction set
> so as not to be bothered by pre-existing licensing agreements with
> AMD, and thus making sure there'd be no competing implementation.  The
> IA64 architecture was quite complex, and there are reasons to believe
> that complexity was seen as a virtue (makes it easier to get more
> patents and keep competitors out).

HP then also poured additional stuff into the architecture to make
migration from PA-RISC easier. I imagine this also made stuff vastly
more complex.

>> But years passed and the i386 architecture got better and better,
>> including stuff like MMX, SSE and AVX was incorporated, IA64 couldn't
>> really keep up.

> The IA64 architecture was a resounding success in one area tho: it
> killed most of the competition that was coming from "above" (at least
> DEC's Alpha, SGI's MIPS, HP's PA, and it likely sped up the demise of
> Sun's SPARC, I don't think it had much impact on POWER or PowerPC,
> OTOH) and thus helped open up the server (and supercomputer) market
> for Intel (and AMD).

I think, IBM is big enough and old enough and established enough with
POWER that a "young whippersnapper" like Intel is no real danger to them
in their own enclosed Mainframe walled garden. I believe Apple moving
away from PowerPC did more damage to IBMs aspirations in that market.

For the others: they where either on board from the start (like HP),
where already dead (like DEC/Compaq) or slipping into the embedded
market (like MIPS).

And SPARC: after being bought by Oracle, the end was more or less
directly clear.

>> Dnd when AMD then presented their AMD64 architecture that could run
>> legacy 8bit/16bit/32bit code as fast as the new code, allowing for a
>> smooth transition, the nickname "Itanic" for IA64 became true: It had
>> been dead on arrival.

> To make matters worse, the IA64 arrived very late to the market (IIRC
> something like 3 years later than planned).

Indeed. The German computer magazine c't had many interesting articles
about the IA64 architecture and also quite early painted its dark
future, because of ever slipping sales figures, performance problems,
the failure to deliver on made promises and the increasing pressure of
the i386/amd64 architectures.

Grüße,
S°

-- 
Sigmentation fault. Core dumped.

[OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Stefan Monnier]

> IA64 (Itanium) was completely incompatible with the installed i386 base.
> The first CPUs had a (very slow) compatibility layer, assisted by
> software, so you could run your "legacy" 16bit/32bit applications.

The original plan/claims was that the support for legacy i386
application would be "just as fast".  This never materialized
(unsurprisingly: it's easy to make a CPU that can run efficiency several
slightly different instruction sets (ISA), like your average amd64 CPU which
can run applications using the amd64 ISA, the i386 ISA, the 80286 ISA
or the 8086 ISA, more or less; but it's much harder to make a CPU that
can run efficiently very different ISAs).

> Also the CPU was designed so that many complexities where delegated into
> the compiler to create the most optimal code but the compilers at the
> time where not up to the task, greatly hampering the new architecture.

More specifically, it depended on solving problems against which
compiler writers had banged their heads for several decades already (and
it is still on going).  Worse: it was based on "old new ideas", IOW it
as trying to solve the problems that were already started to disappear
but was set o bump into new problems that were already starting
to appear.

The name Itanic came from the fact that it seemed likely (even quite
early on, meaning a long time before the name "Itanium" was announced)
to several (most?) knowledgeable CPU designers to lead to a monumental
failure ;-)

> Note: when IA64 was designed (starting in 1994 at HP) we where nowhere
> near the limits of the 32bit i386 architecture with RAM and frequency,
> so it made sense, somewhat.

Indeed.  Also, they wanted to move away from the i386 instruction set so
as not to be bothered by pre-existing licensing agreements with AMD, and
thus making sure there'd be no competing implementation.  The IA64
architecture was quite complex, and there are reasons to believe that
complexity was seen as a virtue (makes it easier to get more patents and
keep competitors out).

> But years passed and the i386 architecture got better and better,
> including stuff like MMX, SSE and AVX was incorporated, IA64 couldn't
> really keep up.

The IA64 architecture was a resounding success in one area tho: it
killed most of the competition that was coming from "above" (at least
DEC's Alpha, SGI's MIPS, HP's PA, and it likely sped up the demise of
Sun's SPARC, I don't think it had much impact on POWER or PowerPC, OTOH)
and thus helped open up the server (and supercomputer) market for Intel
(and AMD).

> Dnd when AMD then presented their AMD64 architecture that could run
> legacy 8bit/16bit/32bit code as fast as the new code, allowing for a
> smooth transition, the nickname "Itanic" for IA64 became true: It had
> been dead on arrival.

To make matters worse, the IA64 arrived very late to the market (IIRC
something like 3 years later than planned).


Stefan

[OFFTOPIC] Re: [?] Why should Distros be called as i386 for a 32-bit PC, and as amd64 for a 64-bit PC, when Intel Core PCs are also 64bit systems

[Stefan Monnier]

> Well, nearly. Itanium Merced was 2001 [1] (althoug you wouldn't buy
> /that/ as a private person), DEC Alpha was even 1992 [2];

FWIW And MIPS was there even a bit earlier with their R4000 (tho the
software support for it only appeared some years later: they first
wanted to have an installed base to which to deploy the software), which
I believe was the first 64bit microprocessor.

IIRC the claim back then was that adding 64bit support to the R4000 was
rather cheap (it increased the die area by a few percents only, and
64bit adds were still fast enough not to slow down the overall chip's
frequency).

The same must have been true for the Opterons (except that the increase
in die area much have been even much smaller since the CPU itself had
become a much smaller part of the overall die because of the
incorporation of things like the memory controller and the L1 and L2
caches).

So it was a great move on the part of AMD: cheap to implement but with
an enormous marketing impact.


Stefan