Apple Plans to Announce Move to Its Own Mac Chips at WWDC [ARM, TSMC, 5nm, A14]

[Arlen Holder]

Dateline today:
We knew it was coming; where this is when it will be announced...

All verbatim below given apologists incessantly claim all facts are lies.
"Apple Inc. is preparing to announce a shift to its own main
processors in Mac computers, replacing chips from Intel Corp."

"Apple is using technology licensed from Arm Ltd."

"TSMC will build the new Mac processors using a 5-nanometer..."

Apple Plans to Announce Move to Its Own Mac Chips at WWDC
<https://www.bloomberg.com/news/articles/2020-06-09/apple-plans-to-announce-move-to-its-own-mac-chips-at-wwdc>
"future Macs will still run the macOS operating system rather
than the iOS software "

"Apple's chip-development group, led by Johny Srouji, decided to make
the switch after Intel¢s annual chip performance gains slowed. "

"The company is working on at least three of its own Mac processors...
with the first based on the A14 processor in the next iPhone"
--
Bringing TRUTH to Apple ng by consistent application of basic facts.

[Alan Baker]

You realize this is from Bloomberg, right?

And that it's a rumour, right?

[JF Mezei]

On 2020-06-09 14:27, Alan Baker wrote:

> You realize this is from Bloomberg, right?
>
> And that it's a rumour, right?

Mobile Syrup says they got the news from Bloomberg as well. So so far
appears to be single sourced.
> https://mobilesyrup.com/2020/06/09/apple-announce-arm-powered-macs-wwdc/


9to5mac also based n single sourced Bloomberg.
https://9to5mac.com/2020/06/09/apple-arm-mac-starts-wwdc/


When you look at performance improvements of Apple's chip designers on
the Ax line of chips vs what Intel can do on its laptop chips, it seems
to me that it is inevitable that Apple will go ARM.

It left PowerPC for the 8086 because, despite performance advantage on
desktops, because PowerPC lagged for laptops.

If Apple can outperform the 8086 and have better battery life on
laptops, then it stands to gain market share even if it means it won't
sell laptops to those wishing to run windows or linux on it (small
market I would assume)

What is interesting in all of this, in a big picture way is that Intel
saw in the early 1990s the end of the line of the 8086 and started a 64
bit chip project that led to the Itanium replacing the 8086 as "industry
standard".

For Pentium III, the 8086 engineers got "inspiration" from Digital's
Alpha and managed to boost performance a lot. Meanwhile, the IA64
engineers struggled and were very late. By the time the first IA64
machine made first boot, the 8086 already on market outperformed it.
This got worse when AMD forced Intel's hand to upgrade the 8086 to 74
bits and introduced new memory controller that made multi core
performance possible. (that tech was legally inherited from Alpha when
HP donated all the Alpha IP and engineers to Intel).

The irony is that the 8086, instead of dying in the 1990s as predicted,
ended up killing the Itanium chip that was meant to replace it. Now that
the 8086 is finally reaching the end, Intel has no replacement for it,
and it is ARM that will become industry standard.

*IF* Apple makes a Mac-on-ARM announcement this June, watch to see what
happens to Windows and Linux. Could this start a final migration off of
the 8086 onto ARM? If so, Intel would become just another ARM licensee
and it would be up to its engineers to be better than Qualcomm, Apple,
Huawei, Samsung and all the others who make ARM chips. (aka: lots of
competition from people who already have much experience on ARM).

Consider that Intel failed with IA64/Itanium. Intel failed with mobile
modems. Can it succeed as an ARM licensee against serious competitors?

[Arlen Holder]

On Tue, 9 Jun 2020 16:58:16 -0400, JF Mezei wrote:

> On 2020-06-09 14:27, Alan Baker wrote:
>
>> You realize this is from Bloomberg, right?
>>
>> And that it's a rumour, right?
>
>
> Mobile Syrup says they got the news from Bloomberg as well. So so far
> appears to be single sourced.

To JF Mezei (and other adults on this newsgroup),

Thank you for posting as an adult, in your response to Alan Baker.
o I ignore Alan Baker, except when he stalks me around the Internet.

In decades on Usenet, I've only had to plonk two psychopathic stalkers,
Snit and Alan Baker, both of whom are clearly & obviously utter morons,
so I didn't see this idiocy from Alan Baker until you quoted him.

While Alan Baker is an "inane child" (as quoted from VanguardLH over here):
<https://groups.google.com/d/msg/comp.sys.mac.advocacy/miwGEINsoFQ/rjgLMH_cAQAJ>

And while Alan Baker lacks adult "reading comprehension", also noted by
multiple adults in that recent Windows thread, there's nothing wrong with
adults discussing what is reported in the reliable news media, where
Bloomberg, overall, is reliable (IMHO).

Since Alan Baker has the reading comprehension of a child, it's hard to
tell _why_ he even posted what he posted as the entire original post,
including the subject line, was VERBATIM from the cited news report.

Hence, despite the child-like apologist Alan Baker not liking what is
reported in the news, I'm happy to see that you sought to verify the
veracity of what clearly _was_ reported in the news, which, in snippet
form, was clearly:

"Apple Inc. is preparing to announce a shift to its own main

processors in Mac computers, replacing chips from Intel Corp.,
as early as this month at its annual developer conference,
according to people familiar with the plans."

Why does Alan Baker hate any news about Apple that doesn't fit into his
pre-conceived cultist ideas, particularly this simple expectation of news
to be reported this week based on what Bloomberg termed "people familiar
with the plans"? I don't know why.

But I'm glad you, as an adult, responded appropriately to his complaint.
o The conference, AFAIK, should open tomorrow where we'll find out!
--
Apple Apologists like Alan Baker likely have an IQ of around 40 or 50.

[Alan Baker]

On 2020-06-21 7:38 p.m., Arlen Holder wrote:
> On Tue, 9 Jun 2020 16:58:16 -0400, JF Mezei wrote:
>
>> On 2020-06-09 14:27, Alan Baker wrote:
>>
>>> You realize this is from Bloomberg, right?
>>>
>>> And that it's a rumour, right?
>>
>>
>> Mobile Syrup says they got the news from Bloomberg as well. So so far
>> appears to be single sourced.
>
> To JF Mezei (and other adults on this newsgroup),
>
> Thank you for posting as an adult, in your response to Alan Baker.
> o I ignore Alan Baker, except when he stalks me around the Internet.

So this is you "ignoring me", Arlen?

[Arlen Holder]

On Mon, 22 Jun 2020 02:38:27 -0000 (UTC), Arlen Holder wrote:

> "Apple Inc. is preparing to announce a shift to its own main
> processors in Mac computers, replacing chips from Intel Corp.,
> as early as this month at its annual developer conference,
> according to people familiar with the plans."

Hi J.F. Mezei,

The announcement of the plans happened today... for real.

o *Apple launches Arm-based Mac chips*
<https://www.cnet.com/news/wwdc-2020-apple-launches-arm-based-mac-chips-ios-14-ipados-14-macos-big-sur-and-more/>
"the big news was Apple's highly anticipated move to Arm-based chips"

o *Apple gives Macs a brain transplant with new Arm chips*
<https://www.cnet.com/news/apple-gives-macs-a-brain-transplant-with-new-arm-chips/>
"The Mac is transitioning" to ARM-based chips, said Tim Cook.
[Tim Cook calls it "Apple Silicon" but that's just a clever
MARKETING term, for reasons known to Apple and which can be
easily guessed why they don't want to call it "ARM-based chips",
which is what it is.

"Apple demonstrated several important apps already adapted"
--
Never underestimate when Apple spends the most in MARKETING and the least
in R&D % in all of high tech, which tells you what Apple really sells.

[nospam]

In article <rcr21b$luh$1...@news.mixmin.net>, Arlen Holder

<arlen...@newmachine.com> wrote:


>
> o *Apple gives Macs a brain transplant with new Arm chips*

if only you would get a brain transplant.

> <https://www.cnet.com/news/apple-gives-macs-a-brain-transplant-with-new-arm-ch

> ips/>
> "The Mac is transitioning" to ARM-based chips, said Tim Cook.
> [Tim Cook calls it "Apple Silicon" but that's just a clever
> MARKETING term, for reasons known to Apple and which can be
> easily guessed why they don't want to call it "ARM-based chips",
> which is what it is.

it's much more than 'arm-based chips'.

[Alan Baker]

On 2020-06-22 12:50 p.m., Arlen Holder wrote:
> On Mon, 22 Jun 2020 02:38:27 -0000 (UTC), Arlen Holder wrote:
>
>> "Apple Inc. is preparing to announce a shift to its own main
>> processors in Mac computers, replacing chips from Intel Corp.,
>> as early as this month at its annual developer conference,
>> according to people familiar with the plans."
>
> Hi J.F. Mezei,
>
> The announcement of the plans happened today... for real.
>
> o *Apple launches Arm-based Mac chips*
> <https://www.cnet.com/news/wwdc-2020-apple-launches-arm-based-mac-chips-ios-14-ipados-14-macos-big-sur-and-more/>
> "the big news was Apple's highly anticipated move to Arm-based chips"
>
> o *Apple gives Macs a brain transplant with new Arm chips*
> <https://www.cnet.com/news/apple-gives-macs-a-brain-transplant-with-new-arm-chips/>
> "The Mac is transitioning" to ARM-based chips, said Tim Cook.
> [Tim Cook calls it "Apple Silicon" but that's just a clever
> MARKETING term, for reasons known to Apple and which can be
> easily guessed why they don't want to call it "ARM-based chips",
> which is what it is.

In the sense that the original architecture was developed by ARM many
years ago.

Not in any sense does ARM design Apple's CPUs.

[Your Name]

It's the other way around - Apple designs ARM-based chips.

But, yes, ARM did design Apple chips. ARM was started by the UK company
Acorn in 1985, but Apple joined project in 1990 for chips to use in the
Newton PDA.
<https://www.telegraph.co.uk/finance/newsbysector/epic/arm/8243162/History-of-ARM-from-Acorn-to-Apple.html>

[Alan Baker]

"Did"? Of course.

:-)

[JF Mezei]

On 2020-06-22 16:29, nospam wrote:

> it's much more than 'arm-based chips'.

From the compiler's point of view, the code that runs on the CPU is ARM.


What is interersting is that Apple didn't use the word ARM today. And
Apple hasn't finalized how it will name the chips either.


Apple would have normally come out with A14 chip this year (ios that the
right number) in September. So it will be interesting to see how much of
a refresh of the product lime and when it is done to capture christmas
sales.


Was quite surprised the Intel stock didn't go down today. If Apple is
able to produce ARM based desktop chips that perform bettter than
whyevere 8086 Intel can make, then Qualcomm and others will also start
producing desktop ARM chips and yo may start to (finally) see the 8086
being replaced as industry standard archittectiure.

[nospam]

In article <WmbIG.9083$RN....@fx13.iad>, JF Mezei

<jfmezei...@vaxination.ca> wrote:

>
> > it's much more than 'arm-based chips'.
>
> From the compiler's point of view, the code that runs on the CPU is ARM.

*whoosh*

[Alan Baker]

On 2020-06-22 4:43 p.m., JF Mezei wrote:
> On 2020-06-22 16:29, nospam wrote:
>
>> it's much more than 'arm-based chips'.
>
> From the compiler's point of view, the code that runs on the CPU is ARM.

Which doesn't make chips designed by Apple any less "Apple silicon" than
running Intel's instruction set makes AMD's chips.

[JF Mezei]

On 2020-06-22 20:03, Alan Baker wrote:
> On 2020-06-22 4:43 p.m., JF Mezei wrote:
>> On 2020-06-22 16:29, nospam wrote:
>>
>>> it's much more than 'arm-based chips'.
>>
>> From the compiler's point of view, the code that runs on the CPU is ARM.
>
> Which doesn't make chips designed by Apple any less "Apple silicon" than
> running Intel's instruction set makes AMD's chips.

I never implied such. The fact remains that Apple has implemneted the
ARM architecture and as such, the code that runs in the CPU runs the ARM
instruction set, as generated by the compiler. This does not diminiosh
the huge performance boosts that Apple has been able to get from its own
implememntations year after year.


And you're going to have that, but P.A.Semi which Apple acquired to make
the Axx chips and all the huge performance boosts came from Digital. The
head of P.A.Semi was one of the lead desigtners for both StromgARM and
Alpha projects at Digital.

[Alan Baker]

On 2020-06-22 5:20 p.m., JF Mezei wrote:
> On 2020-06-22 20:03, Alan Baker wrote:
>> On 2020-06-22 4:43 p.m., JF Mezei wrote:
>>> On 2020-06-22 16:29, nospam wrote:
>>>
>>>> it's much more than 'arm-based chips'.
>>>
>>> From the compiler's point of view, the code that runs on the CPU is ARM.
>>
>> Which doesn't make chips designed by Apple any less "Apple silicon" than
>> running Intel's instruction set makes AMD's chips.
>
> I never implied such.

Actually... ...you kind of did.

> The fact remains that Apple has implemneted the
> ARM architecture and as such, the code that runs in the CPU runs the ARM
> instruction set, as generated by the compiler. This does not diminiosh
> the huge performance boosts that Apple has been able to get from its own
> implememntations year after year.
>
>
> And you're going to have that, but P.A.Semi which Apple acquired to make
> the Axx chips and all the huge performance boosts came from Digital. The
> head of P.A.Semi was one of the lead desigtners for both StromgARM and
> Alpha projects at Digital.

People change jobs all the time.

The chief designer of Windows NT was the the former designer of VMS

VMS: add one to each character and you get:

WNT: Windows New Technology.

And just like when the copied Mac OS in many things, Microsoft couldn't
even be bothered changing some of the function names from the originals.

<http://www3.sympatico.ca/n.rieck/docs/Windows-NT_is_VMS_re-implemented.html>

[Arlen Holder]

On Mon, 22 Jun 2020 19:43:50 -0400, JF Mezei wrote:

> What is interersting is that Apple didn't use the word ARM today. And
> Apple hasn't finalized how it will name the chips either.

Hi JF Mezei,

I _love_ a logical, sensible, reasonable thinker, with adult cognition.

Again, thank you for _not_ being a sadistic duplicitous cultist like the
Type I apologist nospam often is, and specifically for not being a
psychopathic brain-dead cultist like the Type III apologist Alan Baker is.

IMHO, it's clear that the word "ARM" didn't show up for a very good reason.
o One word... MARKETING

More specifically, IMHO, it's what Apple MARKETING is brilliant at doing:
o *Product Differentiation*

Q: Is your new A14 chip architecture ARM based Tim Cook?
A: No. It's not ARM-based; it's better - it's "_Apple Silicon_(TM)!".
--
HINT: It's arm based, no matter what Apple MARKETING wants you to call it.

[Arlen Holder]

On Mon, 22 Jun 2020 16:29:28 -0400, nospam wrote:

Hi nospam,

> if only you would get a brain transplant.

The only way you _can_ respond to facts you don't like (such as the fact
Apple is transitioning the Mac to ARM-based chips), is to act like a child
(which you just did, again).

As such, you're incapable of an adult discussion, nospam.
o On the Windows ng you & Alan Baker get your head handed to you.

You can only survive on Apple newsgroups...
o Because the bar is so low on Apple ngs for actual facts, IMHO.

> it's much more than 'arm-based chips'.

You don't see any facts - all you _can_ see, is the silly MARKETING.
o You can only see what your cultist mentality allows you to see.

*You are utterly blind to the fact Apple is licensing ARM technology*.
--
Apple apologists are blind to facts because they're a cultist mentality.

[Arlen Holder]

On Tue, 23 Jun 2020 09:47:11 +1200, Your Name wrote:

> It's the other way around - Apple designs ARM-based chips.

While Apple is apparently trying to get us to use the MARKETING term,
"Apple Silicon(TM)", here are more related snippets in the news today...

o *Apple's first ARM-based Mac will be available later this year*
<https://www.theverge.com/2020/6/22/21295475/apple-mac-processors-arm-silicon-chips-wwdc-2020>
"It's a big move that means macOS will support native iOS apps
and macOS apps side by side"

o *Do Apple's new Mac chips mean ARM has won?*
<https://www.engadget.com/apple-arm-upscaled-123033144.html>
"ARM underpins pretty much every smartphone and mobile device
in the world, but in the last 15 years it has struggled to
gain traction in high performance computing."

o *This is Apple's roadmap for moving the first Macs away from Intel*
<https://arstechnica.com/gadgets/2020/06/this-is-apples-roadmap-for-moving-the-first-macs-away-from-intel/>
"Longtime Apple users have been through all this before, with the
transition from PowerPC to Intel and now for Intel x86 to ARM.
All the big platform transition hits are coming back.
The transition to ARM from x86 means that some Mac apps will be native
and some won't. For apps that support both x86 and ARM, Apple is
introducing the "Universal 2" binary that will package both codebases
together. For apps that haven't made the transition to ARM yet,
the Rosetta emulator is back as "Rosetta 2" and will now let x86 apps
run on your ARM Mac, albeit with reduced performance."
--
Bringing TRUTH to Apple newsgroups by consistent application of fact.

[David Sankey]

On 23/06/2020 17:02, Arlen Holder wrote:
> o *Do Apple's new Mac chips mean ARM has won?*
> <https://www.engadget.com/apple-arm-upscaled-123033144.html>
> "ARM underpins pretty much every smartphone and mobile device
> in the world, but in the last 15 years it has struggled to
> gain traction in high performance computing."

Which is why ARM is the processor core in the fastest HPC system in the
world, besting the second place system by a factor 2.8?

<https://www.top500.org>

[Alan Baker]

On 2020-06-23 9:02 a.m., Arlen Holder wrote:
> On Tue, 23 Jun 2020 09:47:11 +1200, Your Name wrote:
>
>> It's the other way around - Apple designs ARM-based chips.
>
> While Apple is apparently trying to get us to use the MARKETING term,
> "Apple Silicon(TM)", here are more related snippets in the news today...

It is Apple silicon, Arlen.

They design the chips.

[Alan Browne]

What nospam was referring to is the particular architecture of Apple's
chips which are structured with specialized CPU's for some tasks. (ie:
some focused on performance and some focused on lower power background,
the neural engine, etc.).

I doubt very much an ARM compiler/assembler would compile Apple's code
for their chips. Apple have surely added, modified and extended
instructions in the machine set. They can "get away with this" because
they don't need to publish this to anyone for compatibility. And before
you go there, this is likely agreed to by ARM as long as Apple keep
their extensions to themselves - even if Apple risk some "collision"
with future official ARM instruction updates that aren't compatible.
With strict configuration control, even this can be handled over the
long haul.

As the Apple chip evolution carries on, the desktop/laptop chip family
will evolve to be more appropriate to those environments (more focus on
performance than power consumption compared to the phones/pads, etc.).

[Alan Browne]

On 2020-06-22 20:20, JF Mezei wrote:
> On 2020-06-22 20:03, Alan Baker wrote:
>> On 2020-06-22 4:43 p.m., JF Mezei wrote:
>>> On 2020-06-22 16:29, nospam wrote:
>>>
>>>> it's much more than 'arm-based chips'.
>>>
>>> From the compiler's point of view, the code that runs on the CPU is ARM.
>>
>> Which doesn't make chips designed by Apple any less "Apple silicon" than
>> running Intel's instruction set makes AMD's chips.
>
> I never implied such. The fact remains that Apple has implemneted the
> ARM architecture and as such, the code that runs in the CPU runs the ARM
> instruction set, as generated by the compiler.

See my other reply. Apple's architecture very likely has extended
instructions and thus "the compiler" and/or assembler has to be Apple's
- not ARM's.

Such would be key to Apple getting the most out of their designs in
their high principle of hardware/software integration.

[Alan Browne]

On 2020-06-22 19:43, JF Mezei wrote:
> On 2020-06-22 16:29, nospam wrote:
>
>> it's much more than 'arm-based chips'.
>
> From the compiler's point of view, the code that runs on the CPU is ARM.

What nospam was referring to is the particular architecture of Apple's
chips which are structured with specialized CPU's for some tasks. (ie:
some focused on performance and some focused on lower power background,
the neural engine, etc.).

I doubt very much an ARM compiler/assembler would compile Apple's code
for their chips. Apple have surely added, modified and extended
instructions in the machine set. They can "get away with this" because
they don't need to publish this to anyone for compatibility. And before
you go there, this is likely agreed to by ARM as long as Apple keep
their extensions to themselves - even if Apple risk some "collision"
with future official ARM instruction updates that aren't compatible.
With strict configuration control, even this can be handled over the
long haul.

As the Apple chip evolution carries on, the desktop/laptop chip family
will evolve to be more appropriate to those environments (more focus on
performance than power consumption compared to the phones/pads, etc.).

<corrects mis-reply to nospam's post>

[nospam]

In article <T_IIG.37273$5_4....@fx40.iad>, Alan Browne

<bitb...@blackhole.com> wrote:

> >>> it's much more than 'arm-based chips'.
> >>
> >> From the compiler's point of view, the code that runs on the CPU is ARM.
>
> What nospam was referring to is the particular architecture of Apple's
> chips which are structured with specialized CPU's for some tasks. (ie:
> some focused on performance and some focused on lower power background,
> the neural engine, etc.).

actually, i was referring to apple making more than just cpus.

they have a wide range of chips and that's only going to get wider.

[Alan Browne]

On 2020-06-24 10:24, nospam wrote:
> In article <T_IIG.37273$5_4....@fx40.iad>, Alan Browne
> <bitb...@blackhole.com> wrote:
>
>>>>> it's much more than 'arm-based chips'.
>>>>
>>>> From the compiler's point of view, the code that runs on the CPU is ARM.
>>
>> What nospam was referring to is the particular architecture of Apple's
>> chips which are structured with specialized CPU's for some tasks. (ie:
>> some focused on performance and some focused on lower power background,
>> the neural engine, etc.).
>
> actually, i was referring to apple making more than just cpus.

That too of course.

>
> they have a wide range of chips and that's only going to get wider.

Indeed. And sorry I replied to you rather than Mezei. Coffee hadn't
kicked in.

[Alan Browne]

On 2020-06-23 13:28, David Sankey wrote:
> On 23/06/2020 17:02, Arlen Holder wrote:
>> o *Do Apple's new Mac chips mean ARM has won?*
>> <https://www.engadget.com/apple-arm-upscaled-123033144.html>
>>    "ARM underpins pretty much every smartphone and mobile device
>>     in the world, but in the last 15 years it has struggled to
>>     gain traction in high performance computing."
>

> Which is why ARM is the processor **** core **** in the fastest HPC system in the

> world, besting the second place system by a factor 2.8?
>
> <https://www.top500.org>

**** Plural. Big plural.

Folding@Home beats the above, by the way, at over 2.4 exaflops back in
mid April ... mainly GPU's and CPU's under Windows, MacOS and Linux. It
would be more, but the researchers are having trouble generating enough
work for the network...

[Arlen Holder]

On Wed, 24 Jun 2020 10:50:29 -0400, Alan Browne wrote:

> Folding@Home beats the above, by the way, at over 2.4 exaflops back in
> mid April ... mainly GPU's and CPU's under Windows, MacOS and Linux. It
> would be more, but the researchers are having trouble generating enough
> work for the network...

Hi Alan Browne,

Any logical reasonable person would ask "Why transition to ARM Macs?"

More and more, as I see posts from you, Alan Browne, I wonder if you're a
Type II apologist, or a normal person, as your post above, and a few others
in this thread, appear to come from a normal person (kudos to you).

Your posts in this thread are completely devoid of apologists' idiocies.

As you reasonably and logically noted with respect to performance, what's
important for an adult discussion is _why_ Apple is "transitioning" to the
ARM Mac from the Intel Mac.

There will always be two sides of the answer to that question:
o What Apple "claims" is the real reason, and,
o What the real reason really is.

I'm not hinting that Apple isn't claiming their real reason, since I don't
know their real reason, but any logical sensible adult would look at both.

Lord knows, Apple claimed the batteries made them do it, telling Congress
that throttling on the iPhone 8 and iPhone X wasn't "as" necessary, and
yet, Apple has throttled more and more iPhones in every iOS since iOS 10
when they made that claim... and Apple admitted they secretly throttled on
purpose to lower the life of the iPhone (and they paid a criminal fine for
that admission of guilt), so my point is that what Apple "says" is not
always the case.

Hence any logical reasonable person would ask "Why transition to ARM Macs?"
--
I'm not sure why but I'll keep my eyes & ears tuned to WHY the ARM Mac,
starting first with Apple MARKETING claims, and then looking about for the
real reason (if it differs, which, unfortunately, often it does).

[David Sankey]

On 24/06/2020 15:50, Alan Browne wrote:
> On 2020-06-23 13:28, David Sankey wrote:
>> On 23/06/2020 17:02, Arlen Holder wrote:
>>> o *Do Apple's new Mac chips mean ARM has won?*
>>> <https://www.engadget.com/apple-arm-upscaled-123033144.html>
>>>    "ARM underpins pretty much every smartphone and mobile device
>>>     in the world, but in the last 15 years it has struggled to
>>>     gain traction in high performance computing."
>>
>> Which is why ARM is the processor **** core **** in the fastest HPC
>> system in the world, besting the second place system by a factor 2.8?
>>
>> <https://www.top500.org>
>
> **** Plural.  Big plural.

HPC has been plural ever since the beginning of Cray in the 1980's with
multiple vector processors, it's only ever been a question of how big -
what you see today is ARM up against GPU.

> Folding@Home beats the above, by the way, at over 2.4 exaflops back in
> mid April ... mainly GPU's and CPU's under Windows, MacOS and Linux.  It
> would be more, but the researchers are having trouble generating enough
> work for the network...

Folding@Home includes the above, certainly the bulk of the academic HPC
and compute farms. The Windows and MacOS tend to be real people, you
see the farms turning up under Linux (hence the disparity in GPU to CPU
numbers, as there are a lot of big CPU farms out there).

D

[JF Mezei]

On 2020-06-24 09:59, Alan Browne wrote:

> What nospam was referring to is the particular architecture of Apple's
> chips which are structured with specialized CPU's for some tasks. (ie:
> some focused on performance and some focused on lower power background,
> the neural engine, etc.).

The gizmos around the CPU, sur as neural engine, secure enclave and what
not are separate peripherals. Similar to the graphics cards. Your
program is compiled to run on the CPU with access to the RAM attached to
the CPU. As such, it supports the ARM instruction set.

You can make calls to various system services that lets you make use of
the peripherals such as neural engine. But what you feed as "data" to
the neural engine may not be ARM code. What you feed Secure ENclare are
IO function codes. (on IOS, when you make a disk IO request, it goes
through the secure enclave since it has the physical link to the disk
and controls encryptioN/decryption).

> I doubt very much an ARM compiler/assembler would compile Apple's code
> for their chips. Apple have surely added, modified and extended
> instructions in the machine set.

Nop. Apple provides OS level subroutines (framework in millenial
terminology) to access those devices. Same with graphics cards. The
standard code compiler doesn't generate binary code that will run on a
graphic card. It is either generated separately for a specific card, or
uses standard language such as openGL or Metal for Apple which is
created separately.

The source code file can contain mixed languages with a pre-processor
that extracts statements for one language and converts them into a
system call followed by arguments containing "data" (what it compiled).
The second compiler (the main one) then seens this as a simple
subroutine call with data arguments, compiles it. At run time, that
system serviec is called and it likely then takes the data argument and
moves it to the device which does whatever it does and returns whatever
it returns.


This is common when writing code that uses a database engine since the
"SELECT" statements are not part of COBOL, C or whatever. But then you
showed earlier you had no clue on how database systems are accessed from
source code.


It should be noted that when Apple moved the Axx chips to 64 bits, it
used the ARM supplied architecture and instruction set. It didn't write
its own. If ARM Inc left some opcodes to be used at implementors'
discretion, than Apple mahy have used them to create now opcodes and
then has to manage the LLVM to make use of that opcode when targetting
an Axx chips that has those new Apple-only opcodes.

Apple cannot willy-nilly use any spare opcode because ARM inc may
decided to use it at a later point.

[nospam]

In article <WxOIG.70658$7vd....@fx35.iad>, JF Mezei

<jfmezei...@vaxination.ca> wrote:

> The gizmos around the CPU, sur as neural engine, secure enclave and what
> not are separate peripherals. Similar to the graphics cards.

no they aren't.

[Arlen Holder]

UPDATE: (one person's suggestion why Apple is creating the ARM Mac)

o An ex-Intel engineer claims to know why Apple went in-house
for the future of Mac processors
<https://www.notebookcheck.net/An-ex-Intel-engineer-claims-to-know-why-Apple-went-in-house-for-the-future-of-Mac-processors.477473.0.html>

"This famously straight-talking source is the company's former principal
engineer, François Piednoel. He claims that Apple found fault with
various aspects of Skylake's architecture to such an extent the OEM
was ultimately driven to start making its own processors instead."
--
"These indications remain the opinions of just one engineer, albeit one
with an in-depth knowledge of Intel and its (currently terminated)
relationship with the Cupertino giant."

[Your Name]

On 2020-06-27 21:57:48 +0000, Bud Frede said:

> Alan Baker <notony...@no.no.no.no> writes:
>>
>> People change jobs all the time.
>>
>> The chief designer of Windows NT was the the former designer of VMS
>

> Cutler was "a" designer of VMS, but not "the" designer of it, former or
> otherwise. He did evidently lead the team creating NT, although the
> Microsofties kind of took the silk purse he had worked on and made a
> sow's ear out of it.

That's what Microsloth does with *everything* they buy out, ideas they
steal from other places, and the extremely rare (almost non-existant)
things they actually create themselves.

Adobe does the same, although not quite as badly.

[Alan Baker]

On 2020-06-27 2:57 p.m., Bud Frede wrote:

> Alan Baker <notony...@no.no.no.no> writes:
>
>> On 2020-06-22 5:20 p.m., JF Mezei wrote:
>>> On 2020-06-22 20:03, Alan Baker wrote:
>>>> On 2020-06-22 4:43 p.m., JF Mezei wrote:
>>>>> On 2020-06-22 16:29, nospam wrote:
>>>>>
>>>>>> it's much more than 'arm-based chips'.
>>>>>
>>>>> From the compiler's point of view, the code that runs on the CPU is ARM.
>>>>
>>>> Which doesn't make chips designed by Apple any less "Apple silicon" than
>>>> running Intel's instruction set makes AMD's chips.
>>>
>>> I never implied such.
>>
>> Actually... ...you kind of did.
>>
>>> The fact remains that Apple has implemneted the
>>> ARM architecture and as such, the code that runs in the CPU runs the ARM
>>> instruction set, as generated by the compiler. This does not diminiosh
>>> the huge performance boosts that Apple has been able to get from its own
>>> implememntations year after year.
>>>
>>>
>>> And you're going to have that, but P.A.Semi which Apple acquired to make
>>> the Axx chips and all the huge performance boosts came from Digital. The
>>> head of P.A.Semi was one of the lead desigtners for both StromgARM and
>>> Alpha projects at Digital.
>>
>> People change jobs all the time.
>>
>> The chief designer of Windows NT was the the former designer of VMS
>>
>

> Cutler was "a" designer of VMS, but not "the" designer of it, former or
> otherwise. He did evidently lead the team creating NT, although the
> Microsofties kind of took the silk purse he had worked on and made a
> sow's ear out of it.

Split another hair...

"In June 1975, Cutler, together with Dick Hustvedt and Peter Lipman,
were appointed the technical project leaders for the software project,
code-named Starlet, to develop a totally new operating system for the
Star family of processors. These two projects were tightly integrated
from the beginning."