• Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 11480 from 2003/5/22
    From: Germany
    >> POWER7

    > ...wasn't really the issue here.

    POWER7 does meet your criterion of being one of the "flavors of *current* PPC".

    >> Yes, a fair comparison ... POWER7

    > Sigh.

    Don't make up quotes. I clearly said that comparing POWER7 and current ARM chips based on performance would be *not* fair. I chose POWER7 in that statement because it's the best performing PPC chip in existence currently. This was an answer to you quoting Jim's statement on ARM chips not being "as powerful as PPCs". I don't know which "PPCs" Jim meant to refer to specifically, so I simply chose to take the most powerful one to illustrate my point that comparing performance alone is not fair.

    > Most people in this community who casually makes comparisons of
    > various CPU platforms, usually does that based on one single point of
    > view; Traditional Desktops. Like Crumb and Jim above.

    You could as well exchange "POWER7" for "PA6T" or "PPC970" in my statement. Would work as well to make my point, albeit not as nicely as their performance edge over current ARMs is smaller than POWER7's of course.

    > when desktop-oriented guys makes casual comparisons about architectures,
    > it all really sums up to what's available from Intel and AMD

    Huh? How could a comparison between current ARM chips and current PPC chips (that's what we (including Crumb and Jim) have been talking about, remember?) include anything "from Intel and AMD"?

    > application CPU [...] which is what ARM has been from the beginning.
    > [...] general desktop arena, [...] ARM hasn't got there yet

    ARM architecture was already in the "general desktop arena", which is the arena it was originally developed for. After all, Acorn, who invented the ARM architecture, was a producer of desktop class computers:

    http://en.wikipedia.org/wiki/Acorn_Archimedes
    http://en.wikipedia.org/wiki/Risc_PC
    http://en.wikipedia.org/wiki/Acorn_A7000

    So if it will play a role in desktop computing in the future it will be a *return* for ARM architecture to its origins, not a first.

    > the PPC camp would probably put forward the 5121e, and the ARM camp
    > could choose from, say, the i.MX51, i.MX53, Tegra 1 or Tegra 2. All of those
    > ARM chips performs much better than the PPC

    A core performance comparison of 800 MHz Tegra 1 (ARM11 core) vs. 400 MHz MPC5121e (e300c4 core) should prove interesting I think. Btw, how's your performance comparison between your 800 MHz Efika MX and an 800 MHz Sam440 doing? ;-)

    > One other of the the target application areas of the mobileGT 5121e is
    > automotive infotainment, but I'll tell you what - many car manufacturers
    > are working with ARM and Android now for exactly this area. Many of
    > Freescale's ARM CPU directly targets applications previously owned by PPC.

    I'll tell you what as well. We (that is mostly Neko and me) had this very discussion almost two years ago here on MorphZone. There I absolutely acknowledged that Freescale was substituting PPC (MobileGT) by ARM (i.MX) in automotive *infotainment*, no argument about that at all:

    https://morph.zone/modules/newbb_plus/viewtopic.php?forum=3&topic_id=6428&start=14

    > ARM most often comes out on top in comparisons (performance, features, etc).

    That would depend on the specific application.

    > Home servers/NAS units are other kinds of consumer applications
    > where ARM would compete (and beat) PPC

    Any details to back up the "beat" part of your claim? As a side note, MIPS is strong in the communications and networking market as well:

    http://www.eetimes.com/electronics-news/4206111/MIPS-vs--Power--Truth-about-comms-market-share ("MIPS’ definition of the communication market includes [...] also WiFi, broadband access points and customer premise equipment where MIPS is traditionally strong.")

    > POWER7 doesn't really fit the picture...

    I think you're just mistaken as to why I decided to mention it.

    >> Freescale ... nonsense roadmap

    > Freescale is obviously at least a year behind the competition in the ARM market

    That's not what I was referring to with my remark on this "nonsense roadmap" thing, and you know it. A 4+ GHz quad-core Cortex-A9 based i.MX63 processor in 2011, which is what Freescale had in their April 2010 i.MX roadmap, would not only put Freescale on par with the competition but far far ahead. Only problem was that this roadmap could already be recognized as pure hubris in June 2010 when I found it.

    > to suggest that Freescale wouldn't have a future planned for
    > their i.MX effort is nonsense

    Yes, that's why I didn't do that.

    > Roadmaps are snapshots of how future plans look at that moment
    > in time, nothing else, and future targets are *moving* targets, so much
    > can change after such a snap shot has been taken, so they are only
    > relevant for a limited amount of time.

    You're trying to be funny here? I mean, really, did you find a 4+ GHz quad-core Cortex-A9 based processor supposed to be coming from Freescale (or anybody else for that matter) in 2011 realistic in mid-2010, considering that ARM Ltd. itself has been specifying the Cortex-A9 at 2 GHz "typical operation" as you well know?

    > Your whole perception of reality seems to be based on links and web references

    No, you're mistaken again. It's just that my perception of Freescale's roadmaps is based on what roadmaps Freescale makes publically available on the web. It's certainly not my fault when roadmaps do not even make sense at the time they're published, is it?

    > but that gives a very poor view of the reality

    As you correctly stated before, roadmaps are essentially about plans for the future, less about current reality. If you say that publically available roadmaps give "a very poor view" of a company's plans for the future then what do you think should someone who doesn't have access to confidential roadmaps take as basis for judging the company's plans for the future that gives a better view than the publically available roadmaps?

    > it's a good thing that at least Genesi isn't relying on linked
    > images on websites to get info

    Yes, of course. Genesi are a Freescale business partner which suggests that they have an NDA with Freescale and thus access to confidential Freescale product roadmaps. So they simply don't have to rely on publically available roadmaps from Freescale. I don't know about you but I don't have an NDA with Freescale so I don't have access to confidential Freescale product roadmaps (and if I had I wouldn't be allowed to discuss them in public anyway) and thus have to base my judgements on publically available roadmaps.

    > they are friends with the people who are making those road maps in the first place

    Oh, "friends", really? That's a strange word in the context of business relations I think. But anyway, I would like Genesi to ask their alleged "friends" what they intended by putting a 4+ GHz quad-core Cortex-A9 based i.MX63 processor for 2011 in their April 2010 i.MX roadmap.

    > so they (and others) have a better source or info...

    Yes, of course. Nobody doubted that.

    > You don't understand why I choose to neglect it, do you?

    I think I do.

    > I choose to neglect it because it's *painfully irrelevant*

    No, they *are* consumer products based on PPC, so they're clearly relevant in the context of "consumer products based on PPC" (which is a category that you chose, not me). If you think that something, albeit existing and fitting your category, for whatever reason lacks significance then you're free to mitigate your claim by using for instance "almost nobody" and "almost no". I surely wouldn't object to that.

    > not because I don't know about it, so there is no need for you to try to "enlighten" me.

    I didn't intend to "enlighten" you. I merely asked you if production had stopped as that wouldn't be so surprising to me (web store listings could as well be remnants). And it would have explained your "nobody" and "no" statements very well.

    > it's quite old now and nothing new based on PPC followed it

    That doesn't change the fact that it's "consumer products based on PPC", in our context at least as long as it's still getting produced. You didn't answer my question regarding production status of those devices so I take it they're still being produced.

    > I'll happily *continue* to neglect it and stick to my statement

    Of course I can't force you to be truthful. But I'll continue to take the opportunity to make annotations or corrections whenever I feel you've swept something under the table or have not been truthful.
  • »13.05.11 - 11:30
    Profile