A lot of things is happening on the ARM front.

When it comes to Freescale, they are already producing their i.MX51 series of CPU's. Genesi has been an acknowledged partner in designing products on that (www.freescale.com/genesi), and Genesi is now involved in designing a multi-purpose motherboard that can be used in several product configurations based on the upcoming i.MX53 (BBRV's blog). This will be the first/entry CPU in what they call the "Premium" segment (which this post is all about). It's based on a higher clocked Cortex A8 and will do full HD. The i.MX63 is also under execution and a year away. It's being based on a 40nm Cortex-A9. A year from that, a Cortes-A9 28nm i.MX68 is proposed, and a year from that a 28nm Cortex A-15 (Eagle) i.MX73 is proposed. All these along with numerous lower power CPU's of course, but clearly the ARM architecture is going into new segments where they haven't been before, well into server and desktop territory.

i.MX Roadmap

Another big player in ARM is nVidia. They are already in the "traditional" ARM mobile territory with their Tegra CPU, but they have now announced "that it plans to build high-performance ARM® based CPU cores, designed to support future products ranging from personal computers and servers to workstations and supercomputers."

Along with this, Microsoft has now officially announced support for ARM in Windows 8 (not "CE").

It sure looks like the ARM platform is gaining strength and growing into new markets.

I hope that ARM would be something the MorphOS Team would consider as a future path, now when all interesting PPC options is about to be covered and it's hitting the end of the road.

[ Edited by takemehomegrandma on 2011/1/11 13:00 ]

[ Edited by takemehomegrandma on 2011/1/11 13:02 ]

The nVidia "Denver" project sounds pretty promising. While I think ppc has still a few cards to play (most notably QorIQ with Altivec) I think ARM will have the brighter future because NVidia tries to go the classic computer way additional to the embedded territory. Also ARM has the current buzz - and it is always good to join the buzzing bandwaggon if you also want your share of the buzz..

One of the most critical questions for a hypothetical MorphOS ARM route would be: Will Denver offer a big endian mode? Not all ARMs do offer this feature - but it is critical if a hypothetical MorphOS ARM should get legacy 68k/ppc compability.

If there is no big endian mode avialable ARM has no benefit over x86.

But the biggest question of it all: Who's gonna provide the team the required resources (time, time and time as well as a little incentve money and a few devices). Let's see what comes after the Apple ppc gear is maximally covered (Powerbook, iMac g5, Powermac g5) - and I don't think this is going to happen before 2012.

As a mere user, I don't really mind what CPU is used as long as there's a clear roadmap for the OS.

I've never owned an ARM desktop machine before though, so that would be fun.

Another question. Does an ARM port have to support MorphOS PPC code?
If so, will an ARM processor be powerful enough to emulate a PPC processor?

Quote:

Do you really think that one has a comparable support/momentum?

Quote:

nVidia is far from being the only player here, and I think most manufacturers will have similar offerings. Like Freescale, for example...

Quote:

Again, "Denver" is far from the only interesting CPU here. And I am absolutely no expert, but AFAIK most ARM CPU's are bi-endian. At least I think that's the case with ARMv7/Cortex, which AFAIK includes all interesting stuff here? Do you mean this is a subject to change?

Quote:

Who does that now?

Quote:

Those are the last remaining PPC machines worth supporting. Beyond those, the way I see it:

Option 1: Do nothing, stay PPC. A slow demise for MorphOS until it totally fades away.

Option 2: Go the ARM route. Supports big endian mode, and does have a bright future in broad spectrum, ranging from small mobile devices to powerful desktop and workstation systems. Would need a new 68k JIT compiler. Newer programs in PPC would need to be ported (if a PPC JIT emulator isn't feasible), meaning bad news if you are using "dead" PPC programs...

Option 3: Go the x86 route. Will be a PITA because of endianness stuff AFAIK. Maybe that could be solved in some way? Or start all over with a clean slate, meaning no backwards compatibility at all, which sounds like no fun.

Quote:

Me neither!

Under any circumstances, this is a long term thing...

Quote:

Yes, I think Nvidia's announcement (along with the announcement of ARM support for Windows 8) is significant too.
ARM has primarily been an embedded processor. Now we see large companies focused on the desktop and server market.

If anything, the only problem I see with a move to ARM is that the processors (of the future) will be more powerful than MorphOS can take advantage of (w/o SMP).

I forgot to mention the current buzz about this introduction is that it might be a 64bit processor.

[ Edited by Jim on 2011/1/11 18:15 ]

@ takemehomegrandma

Well shifting to another ISA is far from trivial. Even if endianess matches I guess the job will not be done with some ARM defines and a plain recompile...
Look to the ammount of work that is actually necessary to "just" support another ppc processor like G5 or how long it takes to support a new gfx card. The resources in butterfly country are pretty limited. That may be sad (because I think generally MorphOS has the power to sustain on the market), but unfortunately it is a fact (And with these limited resources the MorphOS team does really an *incredible* job). Who will write a powerful ppc emulator? What about a 68k emulator? Dropping legacy support - well then I don't see much beenfit in ARM over x86. In fact rather the opposite. But this discussion is pointless anyway.

*If* one day there is need to switch ISA and do a cut I'd suggest it like I very briefly outlined here (no big deal whether ARm or x86 then):
https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=6570&forum=3#66869

But who is going to do it? I cannot, and most users here cannot either. Those who can are pretty loaded already.

And yes, Denver *IS* the buzz thing for ARM. As well as MS support is. The other ARMs are interesting, but they are not going to lead ARM to the desktop. And *the desktop is what I am interested in, and not stupid gear like a tablet or mobile phone.
The Genesi netbook is one of the very few intersting ARM offerings so far for general computing. Pity Genesi is rather a mouse than an elephant. It's not their fault, it is the market power of wintel.
Now speaking of wintel: That is the big news: nVidia with MS can break wintel. Maybe in future we will have Wintel and "Windia".

> ARMv7/Cortex, which AFAIK includes all interesting stuff here?

"Denver" won't sport a (current) Cortex core, that's for sure. And maybe it won't even be ARMv7 ISA (i.e. 32-bit), but that's only a guess fueled by 'ARM64' rumours from last November and the better suitability of a 64-bit ISA with regards to server space in general.

https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=6268&forum=11&post_id=80876#80876

Quote:

Never said it would be.

But at some point it will be necessary to either shift, or "call it a day" and quit.

@Andreas_Wolf

Could very well be true.

> the current buzz about this introduction is that it might be a 64bit processor.

Really? Actually, so far I've encountered almost noone in the IT press presenting the assumption that there might be a connection between the November rumours and the recent "Denver" announcement. So I'd object to the allegation that this is what "the current buzz" is about.

Quote:

Actually, I haven't seen anyone in the press willing to speculate on this design. Only people like you and me who are (in an information vacuum) attempting to speculate as to what would be logical.
As you point out, the current ARM cores are not well suited to server tasks. They might be stretched to fulfill desktop duty, but they're not ideal.
I'd go one step further in suggesting that in order to compete with other platforms ARM must move to 64bit. I don't even think my statement is speculative, its a certainty.

Anyone care to take bets on this one?

[ Edited by Jim on 2011/1/11 22:39 ]

>>> the current buzz about this introduction is that it might be a 64bit processor.

>> Actually, so far I've encountered almost noone in the IT press presenting
>> the assumption that there might be a connection between the November
>> rumours and the recent "Denver" announcement. So I'd object to the
>> allegation that this is what "the current buzz" is about.

> Actually, I haven't seen anyone in the press willing to speculate on this design.

So you retract your previous statement that "the current buzz about this introduction is that it might be a 64bit processor"?

Quote:

No that reflects speculation I've heard from others.
I would have to state that Nvidia would be foolish trying to market a 32bit processor in markets dominated by 64 bit processors.
I wouldn't go as far as offering to eat one if it turns out to be 32bit, but I find the prospect of it being anything other than a 64bit design completely unliklely.

Considering the trends in all other ISAs, do you doubt the prospect yourself?

>> So you retract your previous statement that "the current buzz
>> about this introduction is that it might be a 64bit processor"?

> No that reflects speculation I've heard from others.

Again: I fail to see anything combining "Denver" and 64-bit to an extent that would justify calling it "the current buzz about this introduction". As you don't retract your statement I simply *must* have missed all those (non-press) discussions by those "others" all over the web about the prospect of "Denver" being 64-bit. Maybe you have some links for me? Or is it just that you and me differ regarding the meaning of the word "buzz", i.e. that by "the current buzz" you mean merely you and me and some others here on MorphZone discussing it?

> I find the prospect of it being anything other than a 64bit design
> completely unliklely. [...] do you doubt the prospect yourself?

I think I answered that question already by introducing the 64-bit matter to this discussion myself in the first place but it seems I wasn't clear enough unfortunately. So another try: Yes, I doubt the prospect of "Denver" being anything other than a 64-bit design. I think my guess that "Denver" will be 64-bit is an educated one. Still, I doubt that the "bit-ness" is what "the current buzz" is about regarding "Denver". In fact, almost no web article or forum thread (except here on MorphZone) I came across so far even dealt with that question. And yes, I did a Google search and found nothing to the extent of being "the current buzz", only this:

"Questions burning: will it have a 64 bit ISA? If not, that could be a bit of a problem, especially if its going to run windows 8 on desktops (For many server applications, I actually see this as less of a problem). If it is, is it a as of yet undisclosed 64 bit ARM ISA that was rumored a short while ago?"
http://www.semiaccurate.com/forums/showpost.php?p=92883

Quote:

Precisely. The buzz is that Denver will (according to the announcement) have the *power* to challenge high end x86 and that MS will support this. The thing that it will probably be 64 bit is just a symptom, but not the reason.

This is unusual. We appear to be on the same page on this design, but I think I've still managed to offend you.
Yes, you did bring this up first (on this site) AND most people have not made any broad speculations about this chip.
Only a few that have followed current trends have posted any conjecture, so instead of a "buzz" maybe its more like a murmur.
But it is obvious how other ISAs have evolved (even the PPC), so its not unlikely that Nvidia is developing their hardware along a similar line.

"Questions burning: will it have a 64 bit ISA? If not, that could be a bit of a problem, especially if its going to run windows 8 on desktops (For many server applications, I actually see this as less of a problem). If it is, is it a as of yet undisclosed 64 bit ARM ISA that was rumored a short while ago?"

I missed THAT one, but it follows my own intuition. Win8 is inherently 64bit, any 32bit versions (like current versions of Windows) will only be offered to support legacy users.

But none of this require the careful research and thoughtful consideration you're usually associated with. Its just obvious trends and likely conclusions. Why wouldn't it be obvious to the rest of us?

> We appear to be on the same page on this design

We both find it likely that "Denver" is supposed to be a 64-bit design, yes. You said you were of the opinion that this is what "the current buzz about this introduction" was about. I am not of this opinion. So that's where we're on different pages.

> I think I've still managed to offend you.

I don't feel offended. It's just that I still don't get your "buzz" statement and thus am trying to find out what it's all about.

> Only a few that have followed current trends have posted any conjecture

"Almost nobody" fits reality better I think.

> instead of a "buzz" maybe its more like a murmur.

According to my dictionary that's at least closer to what I think it is than "the current buzz" ;-)

> it is obvious how other ISAs have evolved (even the PPC)

Actually, PPC wasn't that late to the show. Timeline of going 64-bit *and* actually shipping corresponding silicon was as follows:

1991: MIPS
1992: Alpha*
1994: PowerPC
1995: SPARC
1996: PA-RISC
2000: z/Architecture*
2001: Itanium*
2002: SuperH
2003: x86

* never had a 32-bit predecessor/counterpart, i.e. has been 64-bit from the get-go

https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=7001&forum=3&start=301 (last link therein)

> its not unlikely that Nvidia is developing their hardware along a similar line.

Yes, but nVidia don't specify ARM ISA. ARM Ltd. does. So if "Denver" really is supposed to be a 64-bit ARM core/chip then this requires ARM Ltd. to at least have plans for something like an 'ARM64' ISA. That's what I believe is highly likely.

> Why wouldn't it be obvious to the rest of us?

I didn't say it wouldn't. I just expressed that I fail to see any alleged "buzz" regarding that matter.

I originally saw news items mentioning this processor on sites covering Windows related hardware/software. I don't think anyone posting responses ever thought to consider that a 64bit ARM processor would be a new development. They've just grown so used to 64bit processors, that they expect all introductions to have this feature by default.

I didn't post anything on this announcement (until I responded to takemehomegrandma's post) for two reason. First, there are no hard details on this design. And second, if it is designed to compete directly with X86 processor, then some of the features its likely to have go beyond what we need for MorphOS. By that I mean 64bit addressing, no doubt multiple cores, Nvidia graphics on die, and possibly some form of virtualization.

Outside of the conventional PC forums, I haven't seen that much interest in this announcement.
MorphZone is the only Amiga related forum I've seen any postings related to this announcement on. I don't think this is directly related to Genesi's move in focus. Rather, I think it's more a product of our base's broader interest range.

I don't blame you for thinking that there no "buzz" on this topic, but I'm surprised how much there has been on sites focused on Wintel platforms. People really seem to want to see someone take on X86 dominance.

Personally, I just think its cool that you and others on this site have been ahead of the curve by discussing this ISA for so long before this popped up.

You all ought to be patting yourself on the back for this one.

[ Edited by Jim on 2011/1/12 2:41 ]

[ Edited by Jim on 2011/1/12 4:56 ]

Quote:

They may be small, but they are always looking for big friends. Pegatron, the now spun off (?) development company of Asus, was heavily involved in the smartbook/smarttop products, and they have a great relationship with Freescale themselves: http://www.freescale.com/genesi

Quote:

WINTEL -> "WARM"

Quote:

nVidia's Denver is one buzz.

Microsoft supporting ARM is another buzz.

Two separate buzz'es, little to do with each other, but that together put solid strength and momentum behind ARM's future evolution.

In their Press Release (see link above), Microsoft clearly focus on SoC's; the high end Denver development of nVidia isn't even mentioned! Instead the focus is on Tegra, in stores since long time, along with CPU's from other manufacturers. SoC's is what MS are interested in. The very title of the PR is "Microsoft Announces Support of System on a Chip Architectures From Intel, AMD, and ARM for Next Version of Windows".

ARM is mentioned as a company (the one owning and controlling the architecture that everyone is using under license), and then the examples of CPU manufacturers are mentioned in plural: "LAS VEGAS — Jan. 5, 2011 — Microsoft Corp. today announced at 2011 International CES that the next version of Windows will support System on a Chip (SoC) architectures, including ARM-based systems from partners NVIDIA Corp., Qualcomm Inc. and Texas Instruments Inc."

I think this has been coming for a long time now. Even long before Windows 7 was released there were heavy rumors floating around that MS would release it for ARM as well. Even though they obviously waited, it seems there were some truth behind the rumors nevertheless; though they didn't release it, they kept it under development. That's why they can demonstrate it today on existing ARM machines:

"At today’s announcement, Microsoft demonstrated the next version of Windows running on new SoC platforms from ... NVIDIA, Qualcomm and Texas Instruments on ARM architecture."

So it *already* runs on current and existing ARM CPU's from more than one manufacturer, and not just in a simple "experimental boot-up proof of technology" or such, but *really* supported:

"The technology demonstration included Windows client support across a range of scenarios, such as hardware-accelerated graphics and media playback, hardware-accelerated Web browsing with the latest Microsoft Internet Explorer, USB device support, printing and other features customers have come to expect from their computing experience. Microsoft Office running natively on ARM was also shown as a demonstration of the potential of Windows platform capabilities on ARM architecture."

Heck, look for yourself in this cool video:
http://www.youtube.com/watch?v=xKc_XGuvNIk


Microsoft already has the high-end x86 desktop and workstation markets cornered, and that market isn't going anywhere. That's also where all the *existing applications* are. This isn't about that, it's about Microsoft has no access to the light-weight ultra mobile market, with low cost, powerful products that doesn't dry out the battery in very few hours. The Atom didn't do it. Microsoft has been standing by to see Apple release their iPad built around their own A4 ARM CPU, which has been enormously popular and dominated this holiday's sales. I have always said that the ARM would prove to scale upwards much easier than the x86 would scale downwards. Microsoft hasn't even been on the same race track, they have been locked out. Now this is about to change. Sure they will continue supporting new SoC developments Intel. But now also AMD and ARM, and I think there is a good reason for that, and a good decision!

> I don't think anyone posting responses ever thought to consider that
> a 64bit ARM processor would be a new development. They've just
> grown so used to 64bit processors, that they expect all introductions
> to have this feature by default.

I don't concur. The fact that ARM to date is 32-bit only is widely known with most people having at least a slight interest in processor architectures. What I find most unfortunate on this matter is that even the tech media who actually did report on the 'ARM64' rumours back in November didn't attempt to connect the "Denver" announcement to those rumours. At least those tech media and their readers must be aware that 'ARM64' isn't there yet. I mean, tech journalism should be more than to just contentually copy press releases, all the more if all they'd have to do to give a better picture is to make a reference to an older report they did themselves. For whatever reason that didn't happen, so the alleged "buzz" regarding the prospect of "Denver" being 64-bit just wasn't (and isn't) there.

> MorphZone is the only Amiga related forum I've seen
> any postings related to this announcement on.

http://www.amiga.org/forums/showthread.php?t=55984
http://eab.abime.net/showthread.php?t=57097
http://moobunny.dreamhosters.com/cgi/mbthread.pl/amiga/expand/182053?page=2
http://www.commodore-amiga.org/en/forum/5-commodore-amiga-hardware-wishlist-/795-nvidia-project-denver-cpu ;-)

> I don't blame you for thinking that there no "buzz" on this topic, but I'm
> surprised how much there has been on sites focused on Wintel platforms.

I'm afraid now you're mixing up things. I've been only objecting to your use of the wording "the current buzz" with regards to "Denver" being 64-bit (as that's what you were referring to, quote: "the current buzz about this introduction is that it might be a 64bit processor"), not to the "Denver" announcement as such (which is the current buzz, no doubt about that). I thought I made this difference clear all the way. There just is no buzz whatsoever "on sites focused on Wintel platforms" about the question of "Denver" being 64-bit or not.

> Pegatron, the now spun off (?) development company of Asus

3 months ago you didn't need the question mark ;-)

https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=6918&forum=3&post_id=77258#77258

> the focus is on Tegra, in stores since long time

To be more precise, the focus is on the ARMv7-A/Cortex-A based Tegra 2, which hadn't been in stores before September 2010 (the Toshiba AC100 being the first device using it AFAIK). The older, much weaker Tegra processors are based on ARMv6/ARM11.

http://en.wikipedia.org/wiki/Nvidia_Tegra#Specifications

About future Tegra chips:

http://www.engadget.com/2010/09/21/nvidia-ceo-tegra-3-almost-done-tegra-4-on-the-way-expect-a-ne/
http://www.netbooknews.com/9920/video-tegra-3-almost-complete-working-on-tegra-4-expect-a-new-tegra-every-year/
http://www.semiaccurate.com/2010/08/12/tegra-3-tapes-out/

On another note, I dug out a rumour from last August about an alleged future x86 compatible chip by nVidia:

http://www.businessweek.com/news/2010-08-13/nvidia-said-to-take-on-intel-in-tablet-computer-chips.html

How would all that (future Tegra chips, "Denver", mystery x86 compatible chip) go together? The following article may have a speculative answer:

"How is an ARM core related to x86? That is easy, Nvidia is going to use Transmeta-esque code morphing firmware to make the CPU run x86 code. [...] The take home message of this whole story is that Nvidia is building an x86 based on the next generation ARM core."
http://www.semiaccurate.com/2010/08/17/details-emerge-about-nvidias-x86-cpu/

Could that actually be what "Denver" is about, having an ARM ISA based chip that is able to morph x86 code into ARM code for 3rd party application compatibility?

"This new processor stems from a strategic partnership, also announced today, in which NVIDIA has obtained rights to develop its own high performance CPU cores based on ARM's future processor architecture. In addition, NVIDIA licensed ARM's current Cortex™-A15 processor for its future-generation Tegra® mobile processors."

"ARM's future processor architecture" sounds like something secret and yet undisclosed, even from ARM.

It's also distinguished from the latest official developments publicly known from ARM, the Cortex-A15 (which is also in the future).

So it's something different. Whether it's "ARM64" remains to be seen...