> like IBM did with their BlueGene by building a new system
> architecture around an array of simplified PowerPC 440 cores.

Just to annotate, recent BlueGene system is built around 8192 "Power BQC" processors at 1.6 GHz each, which I assume have PPC A2 cores (like PowerEN).

http://www.top500.org/system/11073

> the Cortex-A15 that will over time bring:
> • Speeds beyond 2.5GHz+

ARM says "1.5GHz – 2.5 GHz", not "beyond 2.5GHz+".

http://www.arm.com/products/processors/cortex-a/cortex-a15.php?tab=Performance

> 14,000+ DMIPS

14,000 DMIPS with how many cores at what clock frequency? Without these information that figure is of almost no use.

Hi all, long time no see.

I can answer a point here...

Nvidia have 2 options:
1) Take an existing ARM core and tweak it to run faster.
2) Get an Architecture license and design their own core.

In both cases the ISA is the same. The Architecture licensees are required to implement the full ISA.

That said add-ons are not required, the NEON vector extensions (think AltiVec) are optional. Surprisingly Tegra 2 does not have NEON.

Hope that answers that one.

--

One of the problems the MorphOS team have is the Apple machines they are currently targeting have no documentation. At best they'll be able to look at Linux code but otherwise it's reverse engineering.

For an ARM port they can get the docs for the ARM architecture by going to ARM, filling in a form and downloading them.

That won't give them the full docs to whatever chip they target but it's better than reverse engineering.

Getting docs for other parts is likely to be rather more difficult. Companies are getting very paranoid about patent trolls these days so wont give out technical docs without very good reason. Being closed source and commercial might help though.

OTOH Genesi's contacts with ARM could prove very useful.

--

As for 64 bit, scroll to the bottom of this page.



Cheers.

"so the alleged "buzz" regarding the prospect of "Denver" being 64-bit just wasn't (and isn't) there"

Really?

http://www.semiaccurate.com/forums/showthread.php?p=92886

"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?"

There are LOTS of reference to this speculation spread throughout the web. I appreciate Andreas' forceful arguement to the contrary, but he's wrong. He's not the only person on the planet posting this speculation. It doesn't even require his usual in-depth investigation and thoughtful insight.

For once Andreas, admit YOU are wrong.

There has been previous speculation posted on this topic on the Internet. Its not that original .It's easy to find. And your tirade against my claim that its a buzz on the Internet, its tiresome.

For God's sake, find some other inaccuracy in my statements (after all I'm good for them). But this isn't one of them. People ARE speculating as to whether or not Denver is a 64bit processor.


[ Edited by Jim on 2011/1/13 1:29 ]

[ Edited by Jim on 2011/1/13 1:30 ]

[ Edited by Jim on 2011/1/13 1:32 ]

> Nvidia have 2 options:
> 1) Take an existing ARM core and tweak it to run faster.
> 2) Get an Architecture license and design their own core.

I think it has already been established from the start that nVidia is going for option #2 with "Denver". What isn't clear yet is what still to be revealed ARM ISA version they'll base the chip on.

> Hope that answers that one.

Depends on what (you think) the question was about, actually ;-)

> At best they'll be able to look at Linux code

Or even better from a legal point of view: *BSD code :-)

> For an ARM port they can get the docs for the ARM architecture
> by going to ARM, filling in a form and downloading them. That
> won't give them the full docs to whatever chip they target but it's
> better than reverse engineering.

How is having an ARM ISA version's documentation and trying to port an OS to an ARM SoC with a core based on that ARM ISA version any better than having a Power ISA version's documentation and trying to port an OS to a machine with a discrete PPC CPU based on that Power ISA version? You won't even need to fill in a form to get whatever Power ISA version's docs you want:

http://www.power.org/resources/downloads/

So I don't think there's much reverse engineering involved in porting MorphOS to the PPC970 CPU.

> As for 64 bit, scroll to the bottom of this page.

Thanks for this insightful link. To quote the relevant part for lazy bums: "Hitherto, we've decided it's not been sensible to have 64-bit programs. Extended memory addressing at 40 bits is in the latest Cortex-A15 ... but we haven't had the need for a 64-bit [arithmetic logic unit]."

Seeing how the tenses have been used in this statement I believe it can't be ruled out that a yet to be revealed ARM ISA (which "Denver" is supposed to be based on) will be 64-bit. Any objections? Not being a native English speaker I could have interpreted the specific usage of tenses in a wrong way.

>> the alleged "buzz" regarding the prospect of "Denver"
>> being 64-bit just wasn't (and isn't) there

> Really?
> http://www.semiaccurate.com/forums/showthread.php?p=92886

It's almost funny how you use the single link and quote *I* gave you to prove the "almost" part in my "almost no one" claim against me now. Is there really so much buzz on the web about the prospect of "Denver" being 64-bit that you couldn't come up with an own find?

> There are LOTS of reference to this speculation spread throughout the web.

In the posting in which I gave you the link and quote you re-quoted above I asked you for links regarding this (quote: "Maybe you have some links for me?"). So far you've chosen to ignore that question, and your most recent response doesn't add anything of value to answer that question of mine. Capitalized words like "LOTS" aren't going to convince me.

> He's not the only person on the planet posting this speculation

I never claimed I was. Re-read what I actually wrote. There's a wide range between one single person on the planet discussing a matter and that matter being "the current buzz".

> It doesn't even require his usual in-depth investigation and thoughtful insight.

That's correct but doesn't change the fact that I fail to see "the current buzz" about the prospect of "Denver" being 64-bit.

> For once Andreas, admit YOU are wrong.

I'll do as soon as you've convinced me of me being wrong. Remember, only links will do. That shouldn't be too hard I think. (And btw, why "for once"?)

> There has been previous speculation posted on this topic
> on the Internet. Its not that original

Previous than what? Mind you, speculations about the "Denver" announcement can't be older than the "Denver" announcement itself. Any prior speculations about 'ARM64' (to which *I* gave the links btw) do not include the "Denver" announcement.

> It's easy to find.

Then I'd appreciate a link (and please not the one that I gave you).

> your tirade against my claim that its a buzz on the Internet, its tiresome.

So is your tirade against my claim that there's no "buzz on the Internet" about the prospect of "Denver" being 64-bit.

> find some other inaccuracy in my statements

;-)

> this isn't one of them.

I still think it is.

> People ARE speculating as to whether or not Denver is a 64bit processor.

In which places, except here on MorphZone and on the one link I gave you?

Now you want more links?

Yes I can get them.

I can also provide you with PMs I've been exchanging with others for the last week.

But this has gotten a little ridiculous. You said there were no speculations, I've posted one and I'd invite anyone reading this to go look for others. They're there.

If you insist on hammering on this I'll go dig up others, but that's not something I'd prefer to waste my time doing.

You ought to know me well enough to know that I have no problem admitting when I'm wrong. So why would you doubt me when I tell you that this time I'm not?

I will dig up more references if it will halt this unproductive disagreement.

There are people discussing this. You're one. I talked to others. What you and I define as a 'buzz' is obviously different. But the speculation that Nvidia is readying the first 64bit ARM is not new. Intelligent people like yourself are coming to this conclusion in multiple places world wide.

Whether you like it or not, you're part of the 'buzz'.

Hey, like I've said before, as far as I can tell most people in the X86 market never thought to question whether it would be anything other than 64bit.

You're tenacious Andreas. And this time I think you're dead on, but its NOT an original thought. So if you insist, I can dig up more references.
After all, you denied they were there. I've given you one. You don't seriously think that's the only post, do you?

[ Edited by Jim on 2011/1/13 3:00 ]

[ Edited by Jim on 2011/1/13 3:03 ]

http://www.hpcwire.com/blogs/NVIDIA-ARMs-Itself-for-Heterogeneous-Computing-Future-113025584.html

"As part of this strategy, the company has obtained rights to develop its own NVIDIA-designed high performance CPU cores using ARM's future processor architecture. Presumably this will be based on a future 64-bit implementation of the ARM ISA, given that 64-bit computing is the accepted standard outside of the mobile space."

> Now you want more links?

Not only now. I've been asking you for links on page #1 of this thread already.

> I can also provide you with PMs I've been exchanging with
> others for the last week.

To be honest, I'm more interested in web links. I believe them to be stronger evidence than PMs.

> You said there were no speculations

No, I said there were *almost* no speculations. That "almost" part is essential since *I* gave you the one place *I* was able to find (besides MorphZone) where the prospect of "Denver" being 64-bit was discussed on the web.

> I've posted one

Yes, the single one *I* had given to you before to prove *my* point.

> What you and I define as a 'buzz' is obviously different.

That answer to my question on page #1 (quote: "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?") finally came a long way.

> the speculation that Nvidia is readying the first 64bit ARM is not new.

Yes, even my single link I gave you is about 1 day older than my first speculation here on MorphZone. So no need to lecture me on that.

> Intelligent people like yourself are coming to
> this conclusion in multiple places world wide.

Yes, but as far as I can see not to an extent that justifies calling it "the current buzz" about the "Denver" announcement.

> most people in the X86 market never thought to question whether
> it would be anything other than 64bit.

I still don't concur with that.

> its NOT an original thought.

I never claimed so, quite to the contrary.

> you denied they were there.

No, I said *I* couldn't *find* more. That's a difference.

> I've given you one.

...that I had given you before.

> You don't seriously think that's the only post, do you?

No, of course not. But again: I think there's a wide range between one single person (or two or three) on the planet discussing a matter and that matter being "the current buzz".

> http://www.hpcwire.com/blogs/NVIDIA-ARMs-Itself-for-Heterogeneous-Computing-Future-113025584.html

Ah, finally a link to another article. That's more than I had hoped to get from you ;-) Thanks. Besides MorphZone, now we have one article and one forum post. Do you have some more of this "current buzz", possibly?

Absolutely, and if this didn't continue to be one of the more childish exchanges I've had to deal with I'd continue to list them.

But frankly, I'm not sure its worth my time. We're beginning to look badly to the others wallowing through this post and I'm a little embarrassed (for both of us) and not sure if I should continue to encourage you.

This is the first time you've really disappointed me. For some reason you aren't willing to admit the possibility that others have concurrently come to similar conclusions as you have. And you've latched on to the narrow point you think you have about whether or not there is an active discussion of this on the Internet.
I've already provided you with proof this is being discussed.
I heard the speculation before I saw it posted here.

And I don't know why this upsets you because, unlike many other conclusions you've come to, this on involves no great insight. ARM moving to 64bit has been discussed before. Nvidia announces an ARM design (with no details) aimmed at markets dominated by 64bit processor. What else are people to conclude?

Another no-brainer. While Nvidia license existing cores or obtain a license to design its own?
Obviously the later since they infer in their press release that Denver will be more powerful than the A15 (which they state they'll use in other products).
If they want an ARM processor that outperforms the best current design how else are they going to get it? They have to design it (or pay someone else to design it which seem far less likely).

The speculation that ARM will be used as a basis for a design that interprets X86 code (similar to Transmeta)? No, I can't see that. Did you follow the development of Transmeta processors? Their key flaw was really low performance. I have a contact who was part of the team that designed Transmeta evaluation boards. It was a neat idea that didn't execute well.

Nvidia could just as easily buy VIA. Or merge with VIA. That would give them the right to produce X86 processors. Considering Nvidia history of buying out other companies, this path seems far more likely.

You know, I think this is the first time you've actually managed to offend me. In the past I've appreciated your input, knowledge, and references. Your a sharp, insightful person.
But you're also hard headed, more than a little snide, and unwilling to admit when you're wrong.

I'll continue posting these examples if you want, but I think there are other ways we could both be using our time.

[ Edited by Jim on 2011/1/13 4:49 ]

[ Edited by Jim on 2011/1/13 4:50 ]

>> Do you have some more of this "current buzz", possibly?

> Absolutely

So pleast list them, or at least some of them if they're too many.

> not sure if I should continue to encourage you.

You said you want me to admit I'm wrong. Not giving more than the one single link you gave so far (besides the one I gave you before) surely won't encourage me to do so.

> you aren't willing to admit the possibility that others have
> concurrently come to similar conclusions as you have.

I've told you several times that this is not true, and you know it (if not, re-read what I actually claimed and objected to). I even admitted that others have come to this conclusion *before* me and *I* was the one presenting the first evidence of that. So please, stop this lie.

> you've latched on to the narrow point you think you have about
> whether or not there is an active discussion of this on the Internet.
> I've already provided you with proof this is being discussed.

I never said it's not being discussed but merely objected to your claim that it's "the current buzz" about this announcement and claimed that "almost no one" has been discussing this. To support this point let me outline what we actually have, as far as I can see. A Google search for nvidia+denver+arm results in about 1.1 million reported hits here right now. Oppose to that the 2 (or 3 including MorphZone) occurrences of this discussion/conclusion on the web that I'm aware of right now. Sincerely, what other conclusions than that it's not the current buzz about this announcement and that there's *almost* no one discussing this can I possibly draw?

> I don't know why this upsets you

I'm upset because you ask me to admit being wrong about something which I think I'm not wrong about, as all evidence I have so far supports my point. The sad fact is that you claim to have more evidence but are not willing to give it to me so that I could see that it's really what "the current buzz" is about. And I'm also upset because you continue to lie about what I claimed (see above).

> this on involves no great insight.

To quote myself: That's correct but doesn't change the fact that I fail to see "the current buzz" about the prospect of "Denver" being 64-bit.

> ARM moving to 64bit has been discussed before. Nvidia announces
> an ARM design (with no details) aimmed at markets dominated by
> 64bit processor. What else are people to conclude?

Don't you see that this is exactly what I'm about? The conclusion is very much obvious, as you say yourself, but still there's (as far as I can see) almost no discussion about this compared to the impact the "Denver" announcement created generally. That's what I've been suprised about from the beginning. I think ARM Ltd. officially offering a 64-bit ISA would even be bigger news than just the "Denver" announcement.

> While Nvidia license existing cores or obtain a license to
> design its own? Obviously the later

As you can read in my response to minator this has been established from the get-go. I don't know why you bring this up now in your response to me.

> since they infer in their press release that Denver will be more powerful
> than the A15 (which they state they'll use in other products). If they want
> an ARM processor that outperforms the best current design how else are
> they going to get it? They have to design it (or pay someone else to design
> it which seem far less likely).

I don't think that rationale makes sense. Along with a new ISA version ARM Ltd. has always offered own corresponding cores to license. So if nVidia can license a still undisclosed ARM ISA version then I think they could as well instead license the associated (and as well still undisclosed) core from ARM Ltd.
The real rationale is that nVidia explicitly say in their press release that they obtained an ARM architectural license and are going to design their own ARM core for "Denver", rather than license a ready-made core from ARM Ltd. That's as clear as such statement can get.

> The speculation that ARM will be used as a basis for a design that
> interprets X86 code (similar to Transmeta)? No, I can't see that.

Thanks for addressing this speculation. I'm with you on that one. It's hard to see that such a solution could make sense performance-wise. Even the author of the article circulating that speculation agrees with us on that.

> this is the first time you've actually managed to offend me.

Offending you wasn't my intention. But still, I can't admit that I'm wrong when I sincerely believe I'm not. I hope you understand this stance of mine.
In a balancing way I can say that I too feel offended by your attempt to outright lie about what I claimed.

> you're [...] unwilling to admit when you're wrong.

That's not true. If you want I can give you a dozen links from MorphZone alone where I admitted being wrong. But yes, it required not just capitalized words but real evidence to convince me. I don't think that's too much to ask.

> I'll continue posting these examples if you want

Yes, please do so. It's very much appreciated. I think a total of 11 links (i.e. one 100,000th of the amount of current Google hits) would suffice to convince me of the prospect of "Denver" being 64-bit being "the current buzz" about this announcement. That's a more than a fair assessment for "the current buzz", I believe.

"No, I said there were *almost* no speculations."

Yes you did say that.

You also said ""so the alleged "buzz" regarding the prospect of "Denver" being 64-bit just wasn't (and isn't) there".

And you'd like 11 Goggle links? Probably doable.
The funny thing is, the first would be this thread (you're part of the buzz).
Another thing is I'd have to agree with allowing you to define the term and state the qualifications that would satisfy your definitions. That is a political game I won't play into.

So if it makes you feel better, there is no buzz (except for what we both know is there).

Do I want to waste my time carrying this ridiculous argument further?

Not really, I feel it diminishes both of us (and everyone else is getting such a kick out of using the term buzz). How you feel about hardly influences that.

[ Edited by Jim on 2011/1/13 23:30 ]

>> "No, I said there were *almost* no speculations."

> Yes you did say that.

I'm glad you decided to stop this lie. Thanks for that.

> You also said ""so the alleged "buzz" regarding the prospect
> of "Denver" being 64-bit just wasn't (and isn't) there".

Yes, that's exactly what I said. And I'm still of this opinion.

> you'd like 11 Goggle links?

If you can give a total of 11 hyperlinks (not "Goggle links") to articles or forum threads discussing the matter then I'll admit being wrong, which is what you asked me to do. So you could assume it's not that I really want you to give me those links but rather that I think you should be eager to give them to me ;-) Seriously, I'll happily admit being wrong when being convinced of being wrong. For me it's more important to know the truth than being right.

> Probably doable.

Yes, I think so. I really tried to be fair by "demanding" only one 100,000th of current Google hits.

> the first would be this thread

Yes, and second would be the link to the forum post I gave you, and third would be the link to the article you gave me. Makes only 8 to go.

> I'd have to agree with allowing you to define the term
> and state the qualifications that would satisfy your
> definitions.

Of course, an agreement on the quantification would be necessary to come to a final conclusion. I really tried to be fair in my assessment. If you think that 11 is not an appropriate number of required links to support the claim of something being "the current buzz" about an announcement I'm all ears for your proposal.

> That is a political game

I don't see it that way. It's all about making claims and presenting sufficient supporting evidence. Nothing political about that.

> there is no buzz

Fine. I'm willing to accept your retraction from your "buzz" claim. A pity as you've been mere 8 links off ;-)

> except for what we both know is there

...which in my book is still too little to qualify as "the current buzz" about the "Denver" announcement.

> everyone else is getting such a kick out of using
> the term buzz

Let them have their kick. Besides, it's not only been about the term "buzz" as such but more about the whole phrasing "the current buzz" (as opposed to just "a current buzz").

> How you feel about hardly influences that.

Exactly. Nothing about feelings here, only about claims and hard evidence (or so far rather lack thereof, unfortunately).

Andreas, arguing with you is as tiring as arguing with one of my siblings.
But you're right. No hard feelings.
Besides, it more important that we get back to the topic.
I'm extremely excited about Nvidia picking this up and Microsoft's support can only broaden ARM support.

Er, what exactly are you guys arguing about?

Maybe the answer is in the Nvidia press release:

"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."

> what exactly are you guys arguing about?

Read up from where that argument started:

https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=7675&forum=3&post_id=80969#80969

> Maybe the answer is in the Nvidia press release

No, it's not. If it was there wouldn't have been any argument in the first place.

Totally off the point. Andreas and I just have a valid difference of opinion about the number of people who actually have clued into what this announcement may mean.
Frankly I think he's right that virtually none of the so-called IT professionals in the press have realized what the likely specs of this processor are (or how it represents something new).
Only a few people in forums and blogs have put two and two together and drawn these conclusions.
Our only 'argument' (if you can call it that) is on the numbers. And he's probably right, most people appear to be clueless about how significant this could be.

Just a though that occurred to me while I was out walking the dog.
Microsoft intends to support ARM in Windows 8.
Windows supports virtualization.
Were MorphOS to be ported to ARM it is not inconceivable that in the future we could run MorphOS and Windows together simultaneously on the same computer.

[ Edited by Jim on 2011/1/14 2:47 ]

> Microsoft's support can only broaden ARM support.

"The idea that Microsoft is readying a version of Windows 8 for ARM is troubling."
http://www.amiga.org/forums/showpost.php?p=606510

;-)

What can I say, schizophrenia maybe?
I'm not sure if I trust Microsoft, but they could increase interest in the market.
Damned good quote though.

Quote:

Amazing!
Ehi Jim, whenever and wherever you post something, remember.....Andrea-bot is "watching you"

> Ehi Jim, whenever and wherever you post something,
> remember.....Andrea-bot is "watching you" ;-)

https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=7183&forum=3&post_id=76517#76517 ;-)

Btw, are you Italian? Andrea is a female name here in Germany. Trust me, I'm as manly as a man can be ;-)

Here is a video showing a Genesi ARM computer using one of the amazing Pixel Qi screens that will be optional in future Genesi products:

http://bbrv.blogspot.com/2011/01/new-year-ces-and-lots-more.html

Quote:

I thought that was wolfbot, and not only is he watching...

"Btw, are you Italian? Andrea is a female name here in Germany. Trust me, I'm as manly as a man can be "

but the dude's got some seriously frightening recall.

I remember that exchange - Hey Andreas in the US would be a man's name as well - I don't ever remember him forgetting to and the 's' at the end.

[ Edited by Jim on 2011/1/14 15:03 ]

Quote:
It does sound like fun :)

Do you think the original AmigaOS started with software ? Still they had fun creating it... and people created the first pieces of software also did :)