X1000 and A1222
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 10264 from 2003/5/22
    From: Germany
    >> what about the X5000's x4 slot (the one behind the PCIe-PCIe bridge)? Wouldn't
    >> a SATA3 card work full-speed in it, even if it was somehow limited to x2 speed
    >> by the bridge bottleneck?

    > Would it, as its multiplexed?

    I think so, as I doubt that multiplexing reduces the nominal bandwidth by more than 50%. There are 4 input lanes to the bridge, so using x2 bandwidth for a SATA3 card in the x4 slot leaves another x2 bandwidth for the remaining two x1 slots and the PCI bridge with two PCI slots. This means that using further cards in the two x1 slots (or in one x1 slot and one or two PCI slots) would be within the bandwidth of the PCI controller and the SATA3 card wouldn't be slowed down. Just when the two x1 slots *and* one or two PCI slots are occupied by cards, the 4 input lanes have to be multiplexed to more than 4 output lanes, so bandwidth would likely not be enough to sustain full SATA3 speed anymore.

    > Where'd you find the X1 video cards? The only X1 cards I've seen were
    > AMD FirePro cards and similar designs primarily intended to increase the
    > number of displays

    Yes, FirePro and FireMV. In EU online stores, I find currently available new:
    - FireMV 2250 (RV516) for 56 EUR
    - FireMV 2260 (RV620) for 89 EUR
    - FirePro 2450 (RV620) for 271 EUR
    - FirePro 2270 (RV810) for 113 EUR

    > I just did a quick comparison between the P1022 and the T1042's SerDes I/O lines.

    Regarding P1022:
    The P1022 on the Tabor/A1222 riddles me. A-Eon claims one PCIe x16 slot with x4 bandwidth for Tabor. If we look at the possible SerDes configurations for the P1022, we can see that the most sane configurations for PCIe that allow both SATA and Ethernet are x1 x1 x1 (0x06) and x2 (0x0e). This means the x16 slot could have x2 bandwidth at best, not x4. The only option that provides x4 bandwidth for one PCIe controller would be x4 x2 (0x1c), but this lacks both SATA and Ethernet.
    Even if we assume that assignment of the P1022's 6 SerDes lanes was completely arbitrary and not restricted by predefined configurations, PCIe x4 + 2x SATA + GbE would obviously require 7 SerDes lanes.
    Something seems not right here. Any ideas?

    Edit: mystery solved

    Regarding T1042:
    If we look at the possible SerDes configurations (or there, page 37) for the T1042, we see that the most sane configuration is 0x08 with x4 x1 x1 for PCIe, also providing 2x SATA as well as Ethernet (via "Parallel Port").

    > 20 lanes at 5 GHz of the P5040

    ...of which max. 8 can be assigned to PCIe.

    > the only thing that comes close to the 11,2 is Freescale's T4240RDB

    The T4240RDB has one x8 slot and one x4 slot (both at 5 GHz / PCIe v2), so only 50% of slots, 37.5% of PCIe lanes and 75% of PCIe bandwidth. The T4240QDS, which has up to four x4 slots and thus 100% of slots, has only 50% of PCIe lanes but 125% of PCIe bandwidth.

    > the T4240 has 32 SerDes lanes at 10 GHz.

    ...of which max. 16 can be assigned to PCIe (twelve at 5 GHz / PCIe v2, four at 10 GHz / PCIe v3).
    (Beside, the T4241 is to be preferred.)

    > since they're almost ready, how about the X5000 port and the 11,2 port

    I think the X5000 port *is* ready, or isn't it? :-)

    [ Edited by Andreas_Wolf 15.06.2018 - 13:14 ]
  • »26.09.17 - 02:00
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 10264 from 2003/5/22
    From: Germany
    >>>> What about the X5000's x4 slot (the one behind the PCIe-PCIe bridge)? Wouldn't a
    >>>> normal PCIe x16 card work in it? [...] Wouldn't a SATA3 card work full-speed in it,
    >>>> even if it was somehow limited to x2 speed by the bridge bottleneck?

    >>> X5000's x4 slot is blocked by any video card wider than single slot.

    >> So what? Single-slot graphics cards supported by OS4 are widely available,
    >> so this doesn't render my questions invalid.

    > I was not trying to render your question invalid .

    What were you trying to do then by giving that reply? Where's the connection to my specific questions?

    > 89H12NT12G2
    > https://www.idt.com/document/dst/89hpes12nt12g2-datasheet

    I already linked to the PCIe-PCIe bridge two weeks ago. Can you quote the information from the datasheet that answers my specific questions?

    > PI7C9X111SL
    > https://www.diodes.com/assets/Datasheets/PI7C9X111SL.pdf

    Again, what has the PCIe-PCI bridge to do with my questions?
  • »26.09.17 - 02:11
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 10264 from 2003/5/22
    From: Germany
    > I'd assume this is the PCI-E to PCI-E bridge [...]. I figured it was an IDT component,
    > so this confirms it [...]. [...] Thanks for the component listing. [...] In any case, for a
    > technical junkie like myself, good stuff. Again, big thanks. [...] Its better to provide
    > the page where the download link (for the first component) is provided. [...]
    > Still, really f'ing neat stuff. Thanks Spectre.

    When I linked to this same page two weeks ago, nobody seemed to care ;-)

    > Does Tabor use anything like this first bridge?
    > I would certainly help, considering the limited number of SerDes lanes.

    I don't think so as it has just one PCIe slot and apparently no other on-board components connected to PCIe.
  • »26.09.17 - 02:29
    Profile
  • Order of the Butterfly
    Order of the Butterfly
    Spectre660
    Posts: 220 from 2015/6/30
    I was raising a point based on the the models of video card that Jim talks about using so that reply was really pointed in his direction .

    I am looking at a page from the Cyrus Technical Reference Manual.
    I am not an engineer so somebody correct me if I am going wrong.

    The 1 x PCIex4 ,3 x PCIex1 connect through 4 lanes of the P5020/P5040 via the 89H12NT12G2 PCIex-PCIE switch and the 2 PCI slots connect to the 89H12NT12G2 via the PI7C9X111SL PCIE-PCI so if I understand correctly everything is going to pass through 4 lanes to the cpu .

    What would the impact of say transferring data from a drive attached to a pciex4 sata 3 card to a drive attached to a pciex1 sata 2 card while playing audio from internet radio with a PCI based network card and pciex1 Audio card ?

    If you transfer via the pciex4 to the onboard sata no problem ? or if you do the internet radio and transfer from the pciex1 to the onboard sata also no problem ? .

    Quote:

    Andreas_Wolf wrote:
    >>>> What about the X5000's x4 slot (the one behind the PCIe-PCIe bridge)? Wouldn't a
    >>>> normal PCIe x16 card work in it? [...] Wouldn't a SATA3 card work full-speed in it,
    >>>> even if it was somehow limited to x2 speed by the bridge bottleneck?

    >>> X5000's x4 slot is blocked by any video card wider than single slot.

    >> So what? Single-slot graphics cards supported by OS4 are widely available,
    >> so this doesn't render my questions invalid.

    > I was not trying to render your question invalid .

    What were you trying to do then by giving that reply? Where's the connection to my specific questions?

    > 89H12NT12G2
    > https://www.idt.com/document/dst/89hpes12nt12g2-datasheet

    I already linked to the PCIe-PCIe bridge two weeks ago. Can you quote the information from the datasheet that answers my specific questions?

    > PI7C9X111SL
    > https://www.diodes.com/assets/Datasheets/PI7C9X111SL.pdf

    Again, what has the PCIe-PCI bridge to do with my questions?
  • »26.09.17 - 02:45
    Profile
  • Order of the Butterfly
    Order of the Butterfly
    asrael22
    Posts: 383 from 2014/6/11
    From: Germany
    Quote:

    Jim wrote:
    Quote:

    asrael22 wrote:
    Quote:

    Jim wrote:
    Mmm, since they're almost ready, how about the X5000 port and the 11,2 port, and then let's hit the road running for X64 (Ryzen maybe).
    'Cause its not getting just crazy around here, its getting STUPID.
    And I have some tolerance for crazy, but stupid is just plain irritating.


    Yeah, I'd be for it.
    Just bought a 11,2.


    It takes some work to set it up for Linux.
    Your easiest course if you want AMD graphics is one of those Apple X1900GT everyone keeps saying aren't available (although on any given day I can usually find listings for two or three of them).

    If its the water cooled quad, get a t-handled hex wrench, pull them cpus and rebuild the cooler.
    If not you're probably good for another 12 years if you just blow out the dust. ;-)

    BTW - How much did it set you back?


    It's explicitly meant for MorphOS. If support for it doesn't come with 3.10 I'll sell it, or I'll keep it with macOS Leopard for my daughter.
    Probably I paid a little too much, ~100. But it's it good shape, clean, no scratches, etc.


    Manfred
  • »26.09.17 - 07:49
    Profile
  • Priest of the Order of the Butterfly
    Priest of the Order of the Butterfly
    koszer
    Posts: 862 from 2004/2/8
    From: Poland
    Quote:

    asrael22 wrote:
    It's explicitly meant for MorphOS. If support for it doesn't come with 3.10 I'll sell it, or I'll keep it with macOS Leopard for my daughter.



    From what I know the support won't come with 3.10 (if at all). It's way behind schedule to delay it even more (and supporting the PCIe G5's would mean a further delay for the team).
  • »26.09.17 - 07:55
    Profile
  • Order of the Butterfly
    Order of the Butterfly
    Spectre660
    Posts: 220 from 2015/6/30
    @Andreas_Wolf

    A while ago I tried a pciex16 video card in pciex1 slot via a pciex1 to pciex16 adapter but the X5000 would not boot any of the combinations that I tried. Primary Card in pcix16 slot
    X5000 would not boot even with just the adapter and no video card attached with the primary card in the pciex16 slot.
    Second card in a pci slot with pci to pciex16 adapter would boot. Worked but was slow under linux but crashed while workbench was starting up under AmigaOs 4.1
  • »26.09.17 - 10:59
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 10264 from 2003/5/22
    From: Germany
    >> What were you trying to do then by giving that reply?
    >> Where's the connection to my specific questions?

    > I was raising a point based on the the models of video card that Jim
    > talks about using so that reply was really pointed in his direction .

    Okay, so you hit the wrong "Reply" button and quoted the wrong message twice :-)

    > if I understand correctly everything is going to pass through 4 lanes to the cpu .

    Yes, that's what I wrote in comment #40 of this thread, two weeks ago in that other thread and two months ago in yet another thread :-)

    > What would the impact of say transferring data from a drive attached to a
    > pciex4 sata 3 card to a drive attached to a pciex1 sata 2 card while playing
    > audio from internet radio with a PCI based network card and pciex1 Audio card ?

    As I wrote in comment #40, I *think* that the bridge and the SATA3 card could negotiate to use only 2 of the 4 connected lanes, which would still be 67% more bandwidth than required for full-speed SATA3. In your scenario, this would leave x2 bandwidth for the x1 SATA2 card, the x1 audio card and the x1 PCI bridge (with the NIC). This means that the 4 input lanes would have to be multiplexed to 5 output lanes, so bandwidth would likely not be enough to sustain full speed for all cards anymore.
    I don't know what would happen, but I can imagine that the x2 connection between bridge and SATA3 card would be re-negotiated to x1, resulting in reduction to 83% SATA3 speed, but freeing one more lane so that there are 3 lanes for the other three cards connected by x1.

    > If you transfer via the pciex4 to the onboard sata no problem ? or if you do the
    > internet radio and transfer from the pciex1 to the onboard sata also no problem ? .

    I guess so.
  • »26.09.17 - 11:36
    Profile
  • Order of the Butterfly
    Order of the Butterfly
    Spectre660
    Posts: 220 from 2015/6/30
    Did not see post #40 when I was typing what became post #43.
    Was looking at the reference manual page so took my time before pressing submit.
    Sorry also was not following the bridge info in the earlier part of the thread .
    (Sheltering between Hurricanes Irma ,Jose and Maria. Maria passed 65 miles west of me .)


    Quote:

    Andreas_Wolf wrote:
    >> What were you trying to do then by giving that reply?
    >> Where's the connection to my specific questions?

    > I was raising a point based on the the models of video card that Jim
    > talks about using so that reply was really pointed in his direction .

    Okay, so you hit the wrong "Reply" button and quoted the wrong message twice :-)

    > if I understand correctly everything is going to pass through 4 lanes to the cpu .

    Yes, that's what I wrote in comment #40 of this thread, two weeks ago in that other thread and two months ago in yet another thread :-)

    > What would the impact of say transferring data from a drive attached to a
    > pciex4 sata 3 card to a drive attached to a pciex1 sata 2 card while playing
    > audio from internet radio with a PCI based network card and pciex1 Audio card ?

    As I wrote in comment #40, I *think* that the bridge and the SATA3 card could negotiate to use only 2 of the 4 connected lanes, which would still be 67% more bandwidth than required for full-speed SATA3. In your scenario, this would leave x2 bandwidth for the x1 SATA2 card, the x1 audio card and the x1 PCI bridge (with the NIC). This means that the 4 input lanes would have to be multiplexed to 5 output lanes, so bandwidth would likely not be enough to sustain full speed for all cards anymore.
    I don't know what would happen, but I can imagine that the x2 connection between bridge and SATA3 card would be re-negotiated to x1, resulting in reduction to 83% SATA3 speed, but freeing one more lane so that there are 3 lanes for the other three cards connected by x1.

    > If you transfer via the pciex4 to the onboard sata no problem ? or if you do the
    > internet radio and transfer from the pciex1 to the onboard sata also no problem ? .

    I guess so.
  • »26.09.17 - 12:09
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 10264 from 2003/5/22
    From: Germany
    > I tried a pciex16 video card in pciex1 slot via a pciex1 to pciex16 adapter
    > but the X5000 would not boot any of the combinations that I tried.

    Thanks for the info.

    > Primary Card in pcix16 slot X5000 would not boot even with just the adapter
    > and no video card attached with the primary card in the pciex16 slot.

    That's strange as such adapter should be just a dumb extension of the x1 connector (and the other lanes electrically dead or not even there).
  • »26.09.17 - 12:13
    Profile
  • Order of the Butterfly
    Order of the Butterfly
    asrael22
    Posts: 383 from 2014/6/11
    From: Germany
    Quote:

    koszer wrote:
    Quote:

    asrael22 wrote:
    It's explicitly meant for MorphOS. If support for it doesn't come with 3.10 I'll sell it, or I'll keep it with macOS Leopard for my daughter.



    From what I know the support won't come with 3.10 (if at all). It's way behind schedule to delay it even more (and supporting the PCIe G5's would mean a further delay for the team).


    Yeah. This was a highly speculative task to buy this thing. Probably a bit passion driven in the hopes that 3.10 would add support for it.
  • »26.09.17 - 20:04
    Profile
  • Jim
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Jim
    Posts: 4744 from 2009/1/28
    From: Delaware, USA
    Well, I don't see this happening (G5 support) in the next release alongsidr X5000 support (IF we even get that).
    And 3.11 is likely to be a 'bug fix' release.
    Also, they have not stated that support for the 11,2 was even under consideration.
    I bought mine to run Linux, to have something to compare the X5000 to, because it's about the best PPC system released for single user applications, and on the off chance the support might happen.
    But if it doesn't, I still have a competent PPC Linux system.
    I don't know personally how valid the complaints about the X5000 are, but the expansion bus of the 970MP has a LOT of PCI-E lanes.

    [ Edited by Jim 26.09.2017 - 16:02 ]
    "Never attribute to malice what can more readily explained by incompetence"
  • »26.09.17 - 21:01
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 10264 from 2003/5/22
    From: Germany
    > the 970 isn't really a SOC.

    Yes, it's no SoC at all, not even a very basic one like the PA6T with its integrated memory controller (and Ethernet).

    > Hyper Transport...like AMD uses.

    Yes, and HyperTransport is already used by the CPC925/U3/U3H to connect to the PCI(-X) bridge in the currently supported AGP-based G5 PowerMacs.

    > The only reason I have heard for the delays related to 3.10 revolve around
    > the code that had to be replaced.

    So I guess there's no "IFF" to the X5000 support in MorphOS 3.10 :-)
  • »27.09.17 - 23:17
    Profile
  • Jim
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Jim
    Posts: 4744 from 2009/1/28
    From: Delaware, USA
    Cool, and I used to know who invented Hyper Transport, but my memory isn't as strong as yours (which is why I mentioned AMD, but to the best of my knowledge, they are only a licensee).

    These new APUs will make a great addition to AMD's lineup. Ryzen only just competes with Intel's line up, but AMD's graphics are much better than Intels built in gpus.

    Did I misspell 'IF', wouldn't surprise me. ;-)
    "Never attribute to malice what can more readily explained by incompetence"
  • »27.09.17 - 23:47
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 10264 from 2003/5/22
    From: Germany
    > I used to know who invented Hyper Transport [...] (which is why I mentioned AMD,
    > but to the best of my knowledge, they are only a licensee).

    I'm not sure who invented it, but it was introduced in 2001 by the HyperTransport Consortium "founded in 2001 by Advanced Micro Devices, Alliance Semiconductor, Apple Computer, Broadcom Corporation, Cisco Systems, NVIDIA, PMC-Sierra, Sun Microsystems, and Transmeta".
  • »28.09.17 - 00:26
    Profile
  • Jim
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Jim
    Posts: 4744 from 2009/1/28
    From: Delaware, USA
    Quote:

    Andreas_Wolf wrote:
    > I used to know who invented Hyper Transport [...] (which is why I mentioned AMD,
    > but to the best of my knowledge, they are only a licensee).

    I'm not sure who invented it, but it was introduced in 2001 by the HyperTransport Consortium "founded in 2001 by Advanced Micro Devices, Alliance Semiconductor, Apple Computer, Broadcom Corporation, Cisco Systems, NVIDIA, PMC-Sierra, Sun Microsystems, and Transmeta".


    Then most likely Sun.
    "Never attribute to malice what can more readily explained by incompetence"
  • »28.09.17 - 00:44
    Profile
  • Order of the Butterfly
    Order of the Butterfly
    number6
    Posts: 271 from 2008/8/10
    Quote:

    number6 wrote:
    Quote:

    Andreas_Wolf wrote:
    > in October 1000 new AmigaOne A1222 boards will be released to the community.

    Hasn't this already been postponed to early 2018? They must be 3 years old by then.

    > Charge those users a $100 license fee

    I wouldn't find it fair to charge them more than Sam460 user (49 EUR).


    In July, 2017 interview Trevor said that he was "hopeful" that there would be an announcement
    by the end of 2017. If I can find anything more recent from the source, I'll post it.

    Source

    #6




    I promised to post anything new mentioned on this topic. Although not earth shattering, here it is.

    Trevor's most recent blog seems to basically repeat what he has said in the past:

    Quote:

    Although there is still some work needed to done on audio and Ethernet drivers before the AmigaOne A1222 can be commercially released, the FPU optimization appears to be progressing well.


    Nevertheless, apparently they are taking it to AmiWest and intend to do a demo.

    Quote:

    the plan is to show AmigaOS 4.1 running on the AmigaOne A1222 in a live public demonstration for the first time.


    Source for both quotes

    #6
  • »15.10.17 - 16:01
    Profile
  • Cocoon
    Cocoon
    terminills
    Posts: 57 from 2012/3/12
    Quote:

    Jim wrote:
    Quote:

    Andreas_Wolf wrote:
    > I used to know who invented Hyper Transport [...] (which is why I mentioned AMD,
    > but to the best of my knowledge, they are only a licensee).

    I'm not sure who invented it, but it was introduced in 2001 by the HyperTransport Consortium "founded in 2001 by Advanced Micro Devices, Alliance Semiconductor, Apple Computer, Broadcom Corporation, Cisco Systems, NVIDIA, PMC-Sierra, Sun Microsystems, and Transmeta".


    Then most likely Sun.



    It was originally developed by AMD/DEC.

    https://www.extremetech.com/extreme/51158-api-networks-winds-down-engineering-shifts-to-amd



    [ Edited by terminills 16.10.2017 - 12:23 ]
  • »16.10.17 - 11:24
    Profile
  • Jim
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Jim
    Posts: 4744 from 2009/1/28
    From: Delaware, USA
    Thanks, I knew AMD was a significant player in this, I just didn't realize that they were that big a part of it.
    "Never attribute to malice what can more readily explained by incompetence"
  • »16.10.17 - 15:45
    Profile
  • Jim
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Jim
    Posts: 4744 from 2009/1/28
    From: Delaware, USA
    Hmm, I never would have predicted that.
    Looks like a bad decision on AMD's part as it could damage their future growth in laptops once the Ryzen based APUs hit.

    Then again, US companies have always foresaken long term goals for short term profit.

    And maybe I'll be using more Intel CPUs in the future (which might have happened anyway as I prefer discrete graphics cards).
    "Never attribute to malice what can more readily explained by incompetence"
  • »09.11.17 - 10:34
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 10264 from 2003/5/22
    From: Germany
    > I tried a pciex16 video card in pciex1 slot via a pciex1 to pciex16 adapter
    > but the X5000 would not boot any of the combinations that I tried. Primary
    > Card in pcix16 slot X5000 would not boot even with just the adapter and no
    > video card attached with the primary card in the pciex16 slot. Second card
    > in a pci slot with pci to pciex16 adapter would boot. Worked but was slow
    > under linux but crashed while workbench was starting up under AmigaOs 4.1

    Please allow me to link to the update you gave in another thread:
    http://morph.zone/modules/newbb_plus/viewtopic.php?forum=3&topic_id=12178&start=70
  • »13.12.17 - 19:48
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 10264 from 2003/5/22
    From: Germany
    Addendum:

    > The P1022 on the Tabor/A1222 riddles me. [...] PCIe x4 + 2x SATA + GbE would
    > obviously require 7 SerDes lanes. Something seems not right here. Any ideas?

    Mystery solved, and the solution is foreheadslappingly easy: Just as with the newer T1022/T1042, the P1022 does not require any SerDes lanes for the GbE controller if in RGMII mode.
    What had me confused was the simplified P1022 block diagram which implies that the GbE controllers would only work with SerDes lanes assigned to them. Looking at a more detailed P1022 block diagram reveals that the SerDes lanes of the P1022 are assignable to SGMII-mode GbE, but not to RGMII-mode GbE (which is what is used on A1222/Tabor).
    The block diagrams of the P1022RDK board and the P1022DS board also illustrate this.
    So in conclusion, the 6 SerDes lanes of the P1022 supply the PCIe slot with 4 lanes and the two SATA2 ports with 1 lane each, and GbE is provided by the eTSEC/VeTSEC in RGMII mode.
  • »15.06.18 - 11:59
    Profile