> Probably Apple will be the first introducing ARM CPU on desktop machines

http://en.wikipedia.org/wiki/Iyonix_PC
http://en.wikipedia.org/wiki/A9home

> New Mac Mini? Source for rumor?

http://www.macbidouille.com/news/2014/05/24/rumeur-apple-travaillerait-bien-a-des-mac-arm

http://m.seekingalpha.com/article/2245723-apple-arm-based-macs-are-a-fantasy-for-now

Quote:

> "Buy an ARM MacBook - it runs all your apps, just slower!" is a hard sell.

So very true.
The possibility of ARM (or any other RISC) outperforming x86 in everyday use is in distant future, if it ever happens at all.

>>So very true.

This means apps would run very slow. OS X is much slower operating system compared to MorphOS.

....

I think ARM Mac is vaporware. Even if it exists, MorphOS should not port to the ARM Macs. Perhaps SAM or Efika. The Apple company is not friend to MorphOS.

Efika 5200b is already supported.

@KimmoK

Quote:

I don't understand why some people persist in laughing at this thought?!

Already today (or rather since *last fall* even), Apple has live products *on the market* based on its Cyclone ARM CPU architecture. Go back to page 27 in this very thread and read about it, or at least this article:

http://www.anandtech.com/show/7910/apples-cyclone-microarchitecture-detailed

"Apple didn't build a Krait/Silvermont competitor, it built something much closer to Intel's big cores. At the launch of the iPhone 5s, Apple referred to the A7 as being 'desktop class' - it turns out that wasn't an exaggeration."

The article also suggests that it's somewhat of an overkill for mobile phones and tablets, there simply aren't (yet) any available apps to fully harness all this new CPU power, that this CPU didn't really position itself as a competitor to other current mobile CPU's, it's way ahead. Which is all true.

Also read the posts from JuLieN in that same page, and the scores the CPU's got from that chess engine he tested. Those results are impressive. He also compared it to his Intel Core i7 laptop, and if you remove the differences in number of cores and clock frequency from the equation, what does the result tell you?

But the main point is (and here is where everyone falls short in their thinking), you are using *today's* technology to predict *future* products, which isn't very logical, is it? You are saying: "ARM is too slow [today], so Apple could never use it in desktop/laptop [future]". Even though the A7 CPU is somewhat of a "downgraded" (only 2 cores, less cache, etc) and "underclocked (1.3GHz/1.4GHz) Core ix class" CPU (note, I'm not really saying that the A7 is *comparable* to Core ix, but is closer to it than most other ARM CPU's currently in production, and it's not wrong to call the Cyclone design "desktop class")! Today! Not in a "distant future" as you put it, but *today*!

It takes a couple of years to design a CPU (at least from scratch). The Cyclone implementation we see in current products probably started its development a few years back. And given the technology evolution Apple is maintaining, I'd say they have been initiating new development projects on a yearly basis ever since. The products/CPU's we will see *this* fall has been in development for a while (they are probably in production as we speak), and Apple is most certainly already developing the CPU's for 2015's and 2016's product releases right now as well.

The A7 chips that currently powers the iPhone 5s, iPad Air and iPad Mini 2:

• is 64-bit
• is in many aspects similar to desktop/Core ix class CPU's per design
• has 2 cores
• has 64KB(I)+64KB(D) L1 cache, 1MB L2 cache (shared by the 2 cores), 4MB L3 cache
• runs at 1.3GHz or 1.4GHz
• is made in 28nm
• is completely *passively cooled*

This is what is here today (was developed a few years back).

What if:

• Apple moved it to a 20nm process?
• put 4 cores or 6 cores (or more) on each chip instead of just 2? (2MB/3MB L2 cache)
• Dropped the "passively cooled" requirement in favor of powerful, traditional desktop cooling?
• Clocks it at 3+ GHz or more? 4GHz even?
• (+ all kinds of additional on-chip controllers/accelerators and interfaces of course)

What would you have then? How would JuLieN's Stockfish chess engine benchmark, you think? I believe that one (or two ) of these CPU's in a desktop or laptop could potentially make a perfectly fine foundation for "OS XI"...

Bottom line, what I wanted to say was that Apple already has the Cyclone design (it was developed years ago already), it's clearly a paradigm shift in the ARM context and somewhat of an overkill for "only" mobile applications, and if only a decision is made (or was made a few years ago, we don't know that), I'm quite confident that the Cyclone (or a later derivative or forked design, that by all means could already be finished as well) could be scaled up quite easily to produce true desktop class CPU chips that no-one would sneer at. Apple has everything they need, they have the IP and technology, the competence and they have the money. And they would gain a great deal of obtaining complete control over the entire ecosystem. So I have a little difficult understanding those who laughs at this thought?

Apple has changed ISA before. Twice, actually! The last one went very quick, from the announcement that struck the world (and AFAIK even big parts of Apple employees) in 6th of June 2005, it only took a little over a year to switch out their entire product line:

http://en.wikipedia.org/wiki/Apple's_transition_to_Intel_processors

Very fast indeed. And they did it just fine!





[ Edited by takemehomegrandma 13.08.2014 - 20:16 ]

Quote:

...and the "Efika" also exists in ARM versions, both in the form of a "smartbook", and a "smarttop" ("desktop"). I have them both, good computers! But MorphOS for it doesn't exist of course, so...

I only read "Efika"

@takemehomegrandma

It will be interesting to see. I would be surpriced if apple beats ARM mainstream developers.

As Intel BayTrail is slower than Intel desktop...
from june 2013: "Bay Trail-T, clocked at just 1.1GHz, is around 30% faster than Qualcomm’s Snapdragon 800 clocked at 2.3GHz, the fastest ARM chip on the market."

So, I think Apple must somehow get 2* higher performance out of their core than anybody else in ARM land if they think they have intel desktop caliber core.


More comparissons from y2013:
http://www.notebookcheck.net/SoC-Shootout-x86-vs-ARM.99496.0.html
http://www.computingcompendium.com/p/arm-vs-intel-benchmarks.html


[ Edited by KimmoK 13.08.2014 - 16:22 ]

@Kimmok
Apples 64bit A7 in the iphone is running at only 1.3GHZ and reaches nearly Snapdragon 800@2.3GHZ results in geekbench 2. So the "fastest ARM CPU phrase is based on the megahertz myth".

So if you clock it (the A7) at 2.3ghz it would be at minimum as fast as the BayTrail.

Check Tegra 6. Or wait for A8. All they need to do is to reduce size and raise clock. But of course they will add also cores and more cache and new functions and .

> Cyclone [...] is closer to it than any other ARM

Is it? What about X-Gene?

> Apple has changed ISA before. Twice, actually!

6502 -> m68k -> PPC -> x86(-64) is at least three changes.

@KimmoK

Quote:

What do you mean by "ARM mainstream developers"?

Quote:

What phoenixkonsole said above + the fact that Apple has taken a completely different route with the Cyclone design; while most of the others in the ARM mobile market are migrating to whatever next-gen generic cores from ARM that's above their current offering, adding cores to their chips and pumping the clock frequency, Apple has made their own design, and it's a very *wide* chip (á la desktop) that gets (a lot!) more done in less cycles. That's how it can beat quad core chips clocked at ~2GHz, despite it's only a dual core chip running at a mere 1.3GHz.

Quote:

You look at the A7 implementation of the Cyclone, and seem to think something along the lines of "Apple really tried it all here, and they didn't reach desktop performance in their A7 chip". Which is silly. What they made was a chip with only 2 cores, only running at 1.3GHz (in the iPhone, 1.4GHz in the iPad), still providing a vast overkill of performance that simply won't be saturated by any (or at least very few) of the current iOS apps.

But what you forget here, is that they in the Cyclone architecture have a real desktop CPU design!

I merged together a three diagrams from anandtech.com into one picture here:

Cyclone_Vs_Haswell.png

(The Cyclone's picture isn't as detailed, since not everything is known yet, much hush hush, and it's also not official but something put together by Anand Lal Shimpi of anandtech.com based upon his research of documentation and his own tests)

This illustrates how *wide* this Cyclone architecture really is, compared to Intel's Haswell architecture.

To quote anandtech:

"Conroe was a very wide machine. It brought us the first 4-wide front end of any x86 micro-architecture, meaning it could fetch and decode up to 4 instructions in parallel. We've seen improvements to the front end since Conroe, but the overall machine width hasn't changed - even with Haswell."

So the Haswell front end decoders can output up to 4 micro-ops per clock, the Cyclone does 6!

The Out-of-Order window in Haswell is 192. The Cyclone matches this. The Branch Mispredict Penalty is also the same for both (14-19 cycles).

The Haswell has 8 ports to execution units, the Cyclone has 9!

Besides, the Cyclone has *twice* the L1 and L2 cache sice (counted per core, though I understand that the Cyclone's L2 cache is shared?)

The way I see it, is that while the A7 implementation seen in the iPhone and iPad may not exactly be a desktop chip (and it wasn't supposed to be either!), the Cyclone Architecture that it is built from certainly is! To me the Cyclone looks very potent, it's cool and energy efficient for its performance, and I really think it would be possible to build CPU's with quad cores (or more) and clock them really high (maybe ~4GHz in a 20nm process) with suitable and powerful active cooling, and then you would have a chip that's definitely would play in Intel's desktop league. And again, the design of this architecture started a few years ago already, I doubt that in-house evolution has stopped over at Apple. More is coming. Probably on a yearly basis like before. And like Anand say at the end of his article "Swift and Cyclone were two tocks in a row by Intel's definition, a third in 3 years would be unusual but not impossible". We'll see...

Look, I'm not claiming that Apple *will* make an ISA change, but I really believe they *could do it*, if they really wanted. I think the Cyclone shows this. They haven't built a desktop CPU yet based on it, probably more because they didn't *want to* rather than because they coudln't, but I think it would be possible. And not in a "distant future" either...

Quote:

All about your ARM predictions may be true. We'll definitely see more powerful ARMs in future. But Intel/AMD are also progressing. x64 will not just stay where it is now, too.
And x64 will rule the pc market for a good couple of years. The pc market though is a shrinking one due to tablets and stuff like that, but MorphOS is for pcs and not for tablets or phones (unfortunately maybe, not nonetheless true), hence MorphOS should - once ISa gets swapped) - support the most powerful and popular ISA for pcs. And that is x64.

Quote:

Thank you for the education.
So Apple might really have a future winner ARM chip design for desktop kind of use. (I wonder if they ever release it outside their own products)


Btw. are there blender benchmark results on ARM anywhere?

"ARM mainstream developers"

I just meant that Apple is very tiny in ARM world.

[ Edited by KimmoK 14.08.2014 - 09:23 ]

Btw rumors of arm mac mini show a "4 CPU with 4 core each" machine...
This prototype has 4 quad core CPU to have (at its best) enough power to be called a mac MINI? Why not some thousands mos6510? Call me the day arm will have a CPU that outperforms an Intel one...
In the article I linked there's benchmark that shows how an i5 surface disintegrates all ipads on this earth.
Don't misunderstand me: I love wp8 and android on arm but simply try to decompress a big rar file on it and you'll find you're growing very old (Lumia 820-920 and nexus 7)

[ Edited by rebraist 14.08.2014 - 10:54 ]

> Apple might really have a future winner ARM chip design for desktop kind of use.

As far as I understand, it's more about core (microarchitecture) design than chip design.

> Apple is very tiny in ARM world.

Do you have hard numbers?

> Call me the day arm will have a CPU that outperforms an Intel one...

Current ARM cores are faster than many older Intel x86 cores :-)

Quote:

:D i guess a 486... :D
Btw when time will come devs will choose the best "tool"

Quote:

ARM is coming from a position of under 1W per core. You can grow a lot from there.
x86 can't progress the same amount because they can't increase power. If anything they're going the opposite direction.

Once you put an ARM core in a server you can increase the power quite significantly and these are starting to become available now. You can already buy this one:

http://www.enterprisetech.com/2014/08/12/applied-micro-plots-x-gene-arm-server-future/.

Update:

> there's still LSI with its Axxia line of SoCs. It was PPC-only (PPC476FP) originally
> but got ARM members (Cortex-A15) recently.

The Axxia product line is now with Intel:

http://investors.avagotech.com/phoenix.zhtml?c=203541&p=irol-newsArticle&ID=1958352
http://newsroom.intel.com/community/intel_newsroom/blog/2014/08/13/avago-agrees-to-sell-lsis-axxia-networking-business-and-related-assets-to-intel-for-650-million
http://newsroom.intel.com/community/intel_newsroom/blog/2014/08/13/chip-shot-intel-agrees-to-buy-lsis-axxia-networking-business-from-avago
https://communities.intel.com/community/itpeernetwork/datastack/blog/2014/08/13/intel-agrees-to-acquire-the-axxia-networking-business-from-avago-to-boost-wireless-access-networking-strategy
http://www.linleygroup.com/newsletters/newsletter_detail.php?num=5216
http://investors.avagotech.com/phoenix.zhtml?c=203541&p=irol-newsArticle&ID=1990872

Edit: added more links

[ Edited by Andreas_Wolf 03.01.2015 - 00:11 ]

> Transmeta reloaded then.

More news on that:

http://hothardware.com/News/Nvidias-64bit-Tegra-K1-The-Ghost-of-Transmeta-Rides-Again/
http://www.pcworld.com/article/2463900/nvidia-reveals-pc-like-performance-for-denver-tegra-k1.html
http://techreport.com/news/26906/nvidia-claims-haswell-class-performance-for-denver-cpu-core
http://www.zdnet.com/nvidia-details-denver-for-pc-class-performance-in-mobile-7000032525/
http://www.linleygroup.com/newsletters/newsletter_detail.php?num=5205
http://forwardthinking.pcmag.com/none/326531-oracle-nvidia-arm-take-spotlight-at-hot-chips


Edit: added some more

[ Edited by Andreas_Wolf 26.08.2014 - 11:14 ]

"The Apple A8 is a 64-bit system on a chip (SoC) designed by Apple Inc. It first appeared in the iPhone 6, which was introduced on September 9, 2014. Apple states that it is has 25% better CPU performance and 50% graphics performance while drawing only 50% of the power compared to its predecessor, the Apple A7.

The A8 is manufactured on a 20 nm process by TSMC which is a departure from previous processors for Apples mobile devices which was manufactured by Samsung. It has 1 GB of DDR3 RAM included in the package.

Early benchmarking using the Geekbench application suggests that the processor is dual core, and has a frequency of 1.4 GHz. It also supports Apple's claim of it being 25% faster than the A7."

http://en.wikipedia.org/w/index.php?title=Apple_A8


EDIT:

From AnandTech:

CPU: Enhanced Cyclone (i.e. "Cyclone gen 2") @ 1.4GHz, dual core
GPU: PVR GX6650
RAM: 1GB LPDDR3(?), possibly on-chip

http://www.anandtech.com/show/8514/analyzing-apples-a8-soc-gx6650-more

(Everything is still a bit speculative, not taken out of thin air but educated guesses based on the limited info available at this point. I guess things will clear up a bit some time after the 19th.)

[ Edited by takemehomegrandma 10.09.2014 - 12:43 ]

...and there's more:

http://www.anandtech.com/show/8554/the-iphone-6-review/3

Someone on Realworldtech worked out that the SPEC int result should be 1493.
To put it into perspective, that's just below a 2.5GHz G5.

If that's not enough HiSilicon recently announced a 32 core A57 @2.6GHz.

The 64 bit chips are starting to arrive.

Applied Micro Xgene is apparently shipping.

Nvidia K1 is or is soon shipping and getting some pretty stellar geekbench scores.

Mediatech announced the MT6795 8 core A53 / Mali-T760.

Samsung announced the Exynos 7, Quad A57 and quad A53 + Mali-T760.. The international Note 4 has this chip.

Allwinner have a 64 bit chip on the way but don't seem to have said much about it yet.
They do have a dev board on the way:

http://liliputing.com/2014/10/allwinner-nobel64-is-a-64-bit-dev-board.html


In the mean time they'll sell you a $4 quad core A7.

I would have expected 64 bit to appear at the top end and slowly work it's way down.
Look like they'll appear in the mid to low end stuff very rapidly indeed.