PPC Laptop, something for MorphOS?
  • Jim
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Jim
    Posts: 4977 from 2009/1/28
    From: Delaware, USA
    Quote:

    Zylesea wrote:
    The development pace is breath taking! The project is going on now for abot 6 years now.
    Indeed it is kind of a somewhat charming idea generally, but eventually leading to nowhere. Except burning some funds, which could spend elsewhere yield a probably better outcome.


    Six years? Doesn't feel that long. And its a passion project, doesn't have to make sense.

    Further, I really like the T2080, with eight threads it would make a good match for the new kernel Mark wrote. And its bi-endian (on all cores concurrently, although the code needs to be in separate memory pages), and Altivec is limited to big endian mode.

    The biggest drawback I see is price. This thing is likely to be expensive. A Ryzen based laptop would probably cost less (and run circles around a mid-rangr PPC).

    Finally, the pace of development seems to have picked up.

    They hope to from finish the signal integrity testing done soon, then I assume a few tweaks, and they are ready for a prototype run.
    I expect to see this in 2021 (no telling what quarter though).
    "Never attribute to malice what can more readily explained by incompetence"
  • »11.11.20 - 06:07
    Profile
  • Moderator
    Kronos
    Posts: 2231 from 2003/2/24
    Quote:

    Jim wrote:


    Six years? Doesn't feel that long. And its a passion project, doesn't have to make sense.



    Had to go back to the start of the thread, and yes indeed 2014, if you had asked me I would have said >10 years......

    Quote:

    Further, I really like the T2080, with eight threads it would make a good match for the new kernel Mark wrote. And its bi-endian (on all cores concurrently, although the code needs to be in separate memory pages), and Altivec is limited to big endian mode.



    The T2080 was meeeeh in 2014 but today it's more like someone trying to (re)build an 68k-system with a ColdFire SoC in 2010.

    Quote:



    I expect to see this in 2021 (no telling what quarter though).


    Now add time to actually turn this into consumer grade HW and proper SW-support and it will be >10 years.


    Edith got my Trek mixed up:

    As Leonard McCoy will have said in about 150 years:
    "It's dead Jim"



    [ Edited by Kronos 11.11.2020 - 14:32 ]
  • »11.11.20 - 07:54
    Profile
  • Acolyte of the Butterfly
    Acolyte of the Butterfly
    bennymee
    Posts: 132 from 2004/4/14
    From: Netherlands
    Quote:

    Jim wrote:
    Quote:

    Zylesea wrote:
    The development pace is breath taking! The project is going on now for abot 6 years now.
    Indeed it is kind of a somewhat charming idea generally, but eventually leading to nowhere. Except burning some funds, which could spend elsewhere yield a probably better outcome.


    Six years? Doesn't feel that long. And its a passion project, doesn't have to make sense.

    Further, I really like the T2080, with eight threads it would make a good match for the new kernel Mark wrote. And its bi-endian (on all cores concurrently, although the code needs to be in separate memory pages), and Altivec is limited to big endian mode.

    The biggest drawback I see is price. This thing is likely to be expensive. A Ryzen based laptop would probably cost less (and run circles around a mid-rangr PPC).

    Finally, the pace of development seems to have picked up.

    They hope to from finish the signal integrity testing done soon, then I assume a few tweaks, and they are ready for a prototype run.
    I expect to see this in 2021 (no telling what quarter though).


    I really like to see this hardware succeed, but is it a passion project ?
    Imho a project with passion is something you sleap and dream about and want to succeed asap, no matter what is in your way :)


    If you look at the classic scene, e.g. the Warp accelerators, there are 2 people working on it and producing several designs/cards. In their spare time and the Warp boards are not the easiest board to do with interfacing to the old Amiga's.
    According to the 15 sept PDF file there are 50 volunteers, but it is just a comparison...
  • »11.11.20 - 09:49
    Profile
  • Jim
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Jim
    Posts: 4977 from 2009/1/28
    From: Delaware, USA
    Ten years ago, we still had people looking at the 8610, 8640, and 8641 (those were really dead).
    And Tabor hasn't been released yet and it's based on a less powerful core (personally, to be repetitive, I prefer the e5500 cored T104x series).

    To address Kronos' Star Trek quote, unless they're wearing a red shirt, no one stays dead in Star Trek. Long live Power, be it dead or alive.

    I'm in favor of another PPC/Power based system. There is still some potential there.
    Not as much as Ryzen (heck, nobody is touching that these days), but it would provide a viable way to extend legacy MorphOS.

    On production of the first prototypes, anyone care to wager chocolate or beer?

    To close, I really like the proposed expansion capabilities of his laptop. Replaceable graphics, M2 and PCIe/PCI slots. Memory seems limited though.
    "Never attribute to malice what can more readily explained by incompetence"
  • »11.11.20 - 17:22
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 12058 from 2003/5/22
    From: Germany
    > then I assume a few tweaks, and they are ready for a prototype run.

    Only after collection of another 10500 EUR :-)
  • »11.11.20 - 18:00
    Profile
  • Jim
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Jim
    Posts: 4977 from 2009/1/28
    From: Delaware, USA
    Quote:

    Andreas_Wolf wrote:
    > then I assume a few tweaks, and they are ready for a prototype run.

    Only after collection of another 10500 EUR :-)


    Rather economical, by industry standards. And they will have met the goals originally planned.
    Then you can only hope Acube or Aeon will do a production run.

    Again, done in 2021, any wagers?

    BTW - This project has really caught the attention of the posters on Amigaworld.
    Why not? OS4 has planned multi-core support and doesn't support a laptop yet.
    "Never attribute to malice what can more readily explained by incompetence"
  • »11.11.20 - 18:11
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 12058 from 2003/5/22
    From: Germany
    > Memory seems limited though.

    64 GiB if you can find 32 GiB SO-DIMMs, else 32 GiB, according to the quote in comment #165. That's plenty enough for a portable computer I'd say.
  • »11.11.20 - 18:38
    Profile
  • Moderator
    Kronos
    Posts: 2231 from 2003/2/24
    >BTW - This project has really caught the attention of the posters on Amigaworld.
    >Why not? OS4 has planned multi-core support and doesn't support a laptop yet.

    O.k. lets assume these are available next year (just for fun), it would still need a decade for OS4 support to be actually released....

    As for MorphOS support, given that these things always tend to be done by the same 5 guys, I rather have them working on the AMD support both in CPU and GPU and/or doing a "QBox" (SMP,MP all that stuff)


    >To address Kronos' Star Trek quote, unless they're wearing a red shirt,
    >no one stays dead in Star Trek. Long live Power, be it dead or alive.

    Lots of things (and non red shirt people) died and stayed dead in TOS.

    >I'm in favor of another PPC/Power based system.

    I code in C(++) and LUA, couldn't care less if it was running on PPC or some super-hyped up Z80...

    >There is still some potential there.

    It's dead Jim !!!

    >Not as much as Ryzen (heck, nobody is touching that these days),

    Apple seems to be knocking on the basement door, would be surprised to see them at the Penthouse in a few years.

    >but it would provide a viable way to extend legacy MorphOS.

    "legacy MorphOS" is fine as it is and wouldn't benefit much from the T2080 compared to the G4 i/PowerBooks it already supports.

    "MorphOS NG" (if it ever comes to fruition) has no business supporting dead on arrival HW.
  • »11.11.20 - 18:44
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 12058 from 2003/5/22
    From: Germany
    > they will have met the goals originally planned. Then you
    > can only hope Acube or Aeon will do a production run.

    Only after collection of another 26500 EUR for testing and certification :-)
  • »11.11.20 - 18:44
    Profile
  • Jim
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Jim
    Posts: 4977 from 2009/1/28
    From: Delaware, USA
    Quote:

    Andreas_Wolf wrote:
    > they will have met the goals originally planned. Then you
    > can only hope Acube or Aeon will do a production run.

    Only after collection of another 26500 EUR for testing and certification :-)


    Again, by commercial standards, a bargain.

    And Kronos, "its dead" is an opinion.

    And MorphOS NG is being created (at least the kernel) using PPC code.

    Personally, I care what I'm running on, and I have never really cared for X86 (it just had a performance advantage).
    "Never attribute to malice what can more readily explained by incompetence"
  • »11.11.20 - 21:12
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 12058 from 2003/5/22
    From: Germany
    > Again, by commercial standards, a bargain.

    Why apply commercial standards for a passion project, as you put it? The donation campaign started in mid-2017 and is currently at 47%. At that overall pace, it will be finished in 2024.

    > MorphOS NG is being created (at least the kernel) using PPC code.

    The kernel was originally developed on/for the X5000 before being ported to AMD64, yes, but is this still the case? That's not what I got from bigfoot's posting.
  • »11.11.20 - 21:55
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Zylesea
    Posts: 2053 from 2003/6/4
    Quote:

    Andreas_Wolf schrieb:
    > they will have met the goals originally planned. Then you
    > can only hope Acube or Aeon will do a production run.

    Only after collection of another 26500 EUR for testing and certification :-)


    Now imagine what else you could fund with 26500 EUR. Put that money into Wayfarer development or into porting over to x64....

    You can spent money only once, hence I'd put it rather where it moves most in regard you your interests* . And _I_ doubt this is in some ppc hardware.

    *) as this is morph.zone, I just assume that interest #1 is MorphOS.
    --
    http://via.bckrs.de

    Whenever you're sad just remember the world is 4.543 billion years old and you somehow managed to exist at the same time as David Bowie.
    ...and Matthias , my friend - RIP
  • »11.11.20 - 22:08
    Profile Visit Website
  • Jim
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Jim
    Posts: 4977 from 2009/1/28
    From: Delaware, USA
    Valid argument, but people invest in the bounties they want to sponsor.
    Is there a X64 bounty? I'd have to assume you can donate for Wayfarer.

    Regardless, this is moving forward. The drivers for OS4 probably will take forever, if the ludicrous development pace of Tabor's drivers is any indication.
    But the argument that MorphOS NG had to be developed on X64 hardware is invalid.
    You yourself pointed out that high level coding makes the ISA relatively irrelevant.

    Everything done in MorphOS development has been revolutionary.

    Whatever NG becomes it won't be MorphOS without some level of compatibility (even if code requires tweaks and recompilation).

    I obviously don't support Tabor, but I'm running on an unsupported quad core anyway.
    And I can't help but wonder what it would be like to take full advantage of the multi corr systems we do (or could) support.

    I like alternate ISAs, and for a "dead" ISA, Power still seems to have some life in it. ARM would have been cool, too.
    One of those FPGAs with a PPC or ARM core would have made a interesting basis for a MorphOS platform.

    We don't really need the power of a Ryzen based system.
    I'll be glad to see it happen, but at the same time it will be a reminder of how I had to move from 68K to X86.
    "Never attribute to malice what can more readily explained by incompetence"
  • »12.11.20 - 04:26
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 12058 from 2003/5/22
    From: Germany
    > Is there a X64 bounty?

    You mean a bounty for developing MorphOS for x64? Not that I know of.

    > I'd have to assume you can donate for Wayfarer.

    Indeed.

    > the argument that MorphOS NG had to be developed on X64
    > hardware is invalid.

    Which comment contains this argument?

    > You yourself pointed out that high level coding makes the ISA
    > relatively irrelevant.

    Wasn't that Kronos?

    > Everything done in MorphOS development has been revolutionary.

    Everything? Revolutionary? Seriously?
  • »12.11.20 - 06:35
    Profile
  • Jim
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Jim
    Posts: 4977 from 2009/1/28
    From: Delaware, USA
    Sorry, badly composed, I meant evolutionary not revolutionary. And in a way, my own argument works against me since NG is an evolutionary step. Its just a large step (a leap actually).

    I'm waffling too much here, as I can see both sides of this debate.
    I guess I'm basing this on my own preferences and I'm not involved in the decision.
    "Never attribute to malice what can more readily explained by incompetence"
  • »14.11.20 - 13:49
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 12058 from 2003/5/22
    From: Germany
    Update:

    >> then I assume a few tweaks, and they are ready for a prototype run.

    > Only after collection of another 10500 EUR :-)

    The 10500 EUR for 5 prototypes increased to 12500 EUR for 3 prototypes:

    "While interacting with ACube on the simulation process, we were faced with the fact that the verbal agreement we made on the prototyping costs dated back to mid-2017 and the world went through great changes. Back then, they estimated a total cost of €10.500, consisting of a first € 3000 for the initial equipment, and € 1500 for each prototype motherboard, multiplied by 5 motherboards. However, after detailing and updating all involved costs using today’s market quotations, it appears clear that most of the components costs have increased since then [...]. Take for example the NXP T2080 CPU, since 2017 its price has simply dubled, and most of the other components have increased their price too. We discussed extensively with ACube Systems, the initial equipment is still € 3000, but the final cost of each prototype motherboard has increased to € 3000, doubling the initially estimated price of 4 years ago. Because of this dramatic increase in the production cost we decided to make 3 working prototypes only, that makes € 9000. On top of these we add another € 500 to make a dummy board (not working board), printed with a two layers PCB and all mechanical components correctly mounted. The scope of such a dummy board is to ensure that the working prototypes that will be produced later will mechanically fit in the Slimbook Eclipse. As a result, the ongoing campaign goal will be increased to € 12.500."
    https://www.powerpc-notebook.org/2021/02/pcb-gerber-files-published-updates-on-the-prototypes/
  • »05.03.21 - 20:41
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    takemehomegrandma
    Posts: 2720 from 2003/2/24
    I called vapor on this pipe dream years ago, and as the years has gone by since then the thing seems less and less likely to come through and at the same time it gets less and less relevant. PPC is dead, chances are the CPU isn’t even manufactured anymore and that the price raise is due to that. This is a stillborn.
    MorphOS is Amiga done right! :-)
    MorphOS NG will be AROS done right! :-)
  • »05.03.21 - 22:14
    Profile
  • K-L
  • Cocoon
    Cocoon
    K-L
    Posts: 42 from 2020/11/17
    From: Lyon, France
    @takemehomegrandma

    Your opinion.
    PowerMac G5 2,7 Ghz / Radeon 9650 / MorphOS 3.15
    AmigaOne X1000 (unused ATM)
  • »06.03.21 - 04:48
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    takemehomegrandma
    Posts: 2720 from 2003/2/24
    Quote:

    K-L wrote:
    @takemehomegrandma

    Your opinion.


    Of course. But where do *you* draw the line of vaporware? This started almost seven years ago, more than two “computer generations” ago, and still they haven’t even managed to produce even a prototype? Not to mention no plan, infra structure or financials to reach industrialization and commercialization of the project. The enormous amounts of cash needed to be paid up-front *months* ahead of production start for components etc. Will there be another bounty for that? It would be humongous compared to this little one they are struggling to fulfill. And *should* that happen (which it won’t) you will end up with a laptop product that’s completely unsellable in real numbers due to ridiculous price and poor performance compared to *anything* available on the laptop market when it comes out. And then (many years from now) you would have to try to convince OS and driver developers to put in lots of time (thereby money) to support a product that will sell in double (or single!) digit numbers. Designing and building HW is the quick, cheap and easy part of the equation after all. It’s the software side that is difficult.

    [ Edited by takemehomegrandma 06.03.2021 - 07:08 ]
    MorphOS is Amiga done right! :-)
    MorphOS NG will be AROS done right! :-)
  • »06.03.21 - 06:06
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Zylesea
    Posts: 2053 from 2003/6/4
    Quote:

    K-L schrieb:
    @takemehomegrandma

    Your opinion.


    Of course only an opinion. But this project is just a money grave. In fact I think it's even counterproductive and I encourage all ppl interested in morphos to use the money to Support Software development. The bang for the buck in that Domain is far better.
    --
    http://via.bckrs.de

    Whenever you're sad just remember the world is 4.543 billion years old and you somehow managed to exist at the same time as David Bowie.
    ...and Matthias , my friend - RIP
  • »06.03.21 - 06:33
    Profile Visit Website
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 12058 from 2003/5/22
    From: Germany
    > chances are the CPU isn’t even manufactured anymore
    > and that the price raise is due to that.

    NXP claims T2080 availability until 2027, so I don't see this as the likely reason.
  • »06.03.21 - 10:45
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 12058 from 2003/5/22
    From: Germany
    > you would have to try to convince OS and driver developers

    As far as I understand, the components selection was also driven by their state of Linux driver support. And GNU/Linux is the OS this laptop is primarily aimed at.

    > The enormous amounts of cash needed to be paid up-front
    > *months* ahead of production start for components etc. [...]
    > It would be humongous [...]. [...] building HW is the [...] cheap
    > and easy part of the equation after all.

    ;-)
  • »06.03.21 - 11:20
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    takemehomegrandma
    Posts: 2720 from 2003/2/24
    Quote:

    Andreas_Wolf wrote:
    > chances are the CPU isn’t even manufactured anymore
    > and that the price raise is due to that.

    NXP claims T2080 availability until 2027, so I don't see this as the likely reason.


    Could be a declining stock on a shelf. A limited and declining supply is a logical and common reason for big price increases of old stuff that most certainly was written off a long time ago. Look at the PA6T for example. What other reasons for the price of an old (turning ancient) CPU to *double* is there? A sudden vast increase in demand? Not very likely...
    MorphOS is Amiga done right! :-)
    MorphOS NG will be AROS done right! :-)
  • »06.03.21 - 12:57
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    takemehomegrandma
    Posts: 2720 from 2003/2/24
    Quote:

    Andreas_Wolf wrote:
    > you would have to try to convince OS and driver developers

    As far as I understand, the components selection was also driven by their state of Linux driver support. And GNU/Linux is the OS this laptop is primarily aimed at.


    Tell that to the OS4 users that are obviously waiting for this thing to finish. Software support for the Tabor has “been coming” for almost as long as this futile bounty has been running, and they even had to update the Tabor HW design due to the fact that original components went obsolete/cancelled in the meantime.
    MorphOS is Amiga done right! :-)
    MorphOS NG will be AROS done right! :-)
  • »06.03.21 - 13:05
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 12058 from 2003/5/22
    From: Germany
    >>> you would have to try to convince OS and driver developers

    >> As far as I understand, the components selection was also
    >> driven by their state of Linux driver support. And GNU/Linux
    >> is the OS this laptop is primarily aimed at.

    > Tell that to the OS4 users that are obviously waiting for this
    > thing to finish.

    I'm confident they know about this fact, it's all over the project website after all. If this project will fail, it won't be because of a missing OS4 port or any other missing OS port. GNU/Linux compatibility is a given, as I said, and it's essentially the one OS the project maintainers care about. Everything else will be a bonus.

    > Software support for the Tabor has “been coming” for almost
    > as long as this futile bounty has been running

    Tabor, which has been well supported by Linux for years, is all about OS4. This project is not. It's somewhat weird to see you of all people so fixated on OS4.
  • »06.03.21 - 19:41
    Profile