Poster | Thread |
OlafS25
|  |
Re: New hardware: PPC-Motherboard A1222 "Tabor" by Acube/A-eon Posted on 20-Oct-2015 23:15:16
| | [ #641 ] |
|
|
 |
Elite Member  |
Joined: 12-May-2010 Posts: 6472
From: Unknown | | |
|
| @iggy
I had contact to several of those sort... one said AmigaOS was a huge success compared to Aros and MorphOS. Not uncommon. It makes no sense to discuss because it is religion. |
|
Status: Offline |
|
|
iggy
|  |
Re: New hardware: PPC-Motherboard A1222 "Tabor" by Acube/A-eon Posted on 20-Oct-2015 23:25:20
| | [ #642 ] |
|
|
 |
Super Member  |
Joined: 20-Oct-2010 Posts: 1175
From: Bear, Delaware USA | | |
|
| @OlafS25
Quote:
I had contact to several of those sort... |
Yeah, I've never gotten that sentiment from AROS developers. But then, they have been working at this whole NG thing for longer than anyone else.
And what IS 'real Amiga OS'? I honestly think it might be OS3.1, since everything after is third party.
Hard to believe that so many options would create such divisiveness, but you're right, there is a religious aspect to it. |
|
Status: Offline |
|
|
broadblues
 |  |
Re: New hardware: PPC-Motherboard A1222 "Tabor" by Acube/A-eon Posted on 20-Oct-2015 23:47:14
| | [ #643 ] |
|
|
 |
Amiga Developer Team  |
Joined: 20-Jul-2004 Posts: 4449
From: Portsmouth England | | |
|
| @OlafS25
Quote:
Example is the senseless "custom chips" integrated in X1000, no drivers and no developers support. Parts of the hardware are not supported by drivers, to me it seems the hardware developers build something and A-eon and Hyperion tries to support it somehow instead of designing something that fits and can be supported. Normally when creating something OS and hardware devs should work together. To me a little strange.
|
SOCS come with specific network NICS, usb chipsets, etc. If you choose one you have to support what it provides, or increase cost by duplicating the on chip chipsets with seperate chips, which presumably use up PCI lanes (or whatever the righ word is there). And the 'custom chips' are only senseless til someone does something cool with them and then you will all want one 
_________________ BroadBlues On Blues BroadBlues On Amiga Walker Broad |
|
Status: Offline |
|
|
wawa
|  |
Re: New hardware: PPC-Motherboard A1222 "Tabor" by Acube/A-eon Posted on 21-Oct-2015 0:08:24
| | [ #644 ] |
|
|
 |
Elite Member  |
Joined: 21-Jan-2008 Posts: 6259
From: Unknown | | |
|
| @broadblues
Quote:
And the 'custom chips' are only senseless til someone does something cool with them and then you will all want one |
while i dont believe that the explanation is as simple as olaf suggests, i doubt that blindly implementing "features" just for the sake of that someone accidentally might find some use for them, is a strategy ever succesfully employed by anyone. |
|
Status: Offline |
|
|
CodeSmith
|  |
Re: New hardware: PPC-Motherboard A1222 "Tabor" by Acube/A-eon Posted on 21-Oct-2015 3:19:05
| | [ #645 ] |
|
|
 |
Elite Member  |
Joined: 8-Mar-2003 Posts: 3045
From: USA | | |
|
| @itix
Quote:
itix wrote: @Spectre660
Quote:
Anyway A-Eon now own DvPlayer and Tunenet. Two programs that would be affected by the FPU emulation speed. So special SPE floating point versions of these should not be a problem.
|
There will be difficulties because the system ABI has no concept of SPEFP. But that is just another item on the todo list.
|
Elf's multiple section format lends itself naturally to fat binaries and Hyperion maintains the OS4 port of GCC, so this is not a real problem. DOS.library on a Tabor could look for a .text.spe section in the binary and load that in preference to .text, whereas DOS.library on say a SAM460 would just ignore the .text.spe section and load .text.
|
|
Status: Offline |
|
|
cdimauro
|  |
Re: New hardware: PPC-Motherboard A1222 "Tabor" by Acube/A-eon Posted on 21-Oct-2015 4:10:23
| | [ #646 ] |
|
|
 |
Elite Member  |
Joined: 29-Oct-2012 Posts: 4127
From: Germany | | |
|
| @jaokim
Quote:
jaokim wrote:
From that it seems evident that the cheapest e500mc comes at $71, where the P1022 e500 is at $31.
Looking at the chart, it seems that the P1022 is the only choice in the lower price segment, with GHz in par/slightly above the Sam460. The T1040 adds only $10 to the price, but perhaps it wasn't available at the time, plus that the cost increases with $10000 for 1000 units. |
Which isn't that much. Quote:
Perhaps it was deemed cheaper to implement the FPU-fix-thingie for OS4, and start with the hardware? |
Only if you want to just emulate the FPU, because the emulator is already available.
But, as it was emerged, they are improving the support, and it may well cost more than $10000.
That's without counting the cost of splitting the OS4 platform with new PowerPC SPE support & binaries (see above itix's comment about it).
@OlafS25
Quote:
OlafS25 wrote: @iggy
regarding Paul from Varisys... he has no clue about Amiga and Trevor no clue about hardware so Trevor must trust what Paul or others recommend. Example is the senseless "custom chips" integrated in X1000, no drivers and no developers support. Parts of the hardware are not supported by drivers, to me it seems the hardware developers build something and A-eon and Hyperion tries to support it somehow instead of designing something that fits and can be supported. Normally when creating something OS and hardware devs should work together. To me a little strange. |
Yes, in theory. In practice we see the results of such "collaboration"...
@broadblues
Quote:
broadblues wrote: @OlafS25
And the 'custom chips' are only senseless til someone does something cool with them and then you will all want one 
|
I doubt that all X1000 users are coders. And I doubt that all X1000 coders are interested on working on the XCore, because it can be used only for doing some automation.
The reality is that such chip was integrated only to recall the Amiga's custom chips, because OS4 users love to think that "my machine is a new Amiga", and having something which is called as "custom chip", with a "custom bus" (Xorro), serves perfectly to the selling purpose.
So, just marketing, because that chip isn't custom (and the "bus" is just a collector of the extra XCore lines put on a PCI-e connector) and is light years far away for offloading the CPU (and the o.s. too) from typical workloads, like it happened with the Amigas (where ALL coders used the custom chip either directly or through the o.s. APIs).
@CodeSmith
Quote:
CodeSmith wrote: @itix
Quote:
itix wrote: @Spectre660
There will be difficulties because the system ABI has no concept of SPEFP. But that is just another item on the todo list.
|
Elf's multiple section format lends itself naturally to fat binaries and Hyperion maintains the OS4 port of GCC, so this is not a real problem. DOS.library on a Tabor could look for a .text.spe section in the binary and load that in preference to .text, whereas DOS.library on say a SAM460 would just ignore the .text.spe section and load .text.
|
That can only solve the problem of having duplicated code both for regular PowerPC and PowerPC/SPE, but doesn't solve the ABI problem which itix exposed. |
|
Status: Offline |
|
|
Flasheart
|  |
Re: New hardware: PPC-Motherboard A1222 "Tabor" by Acube/A-eon Posted on 21-Oct-2015 6:04:36
| | [ #647 ] |
|
|
 |
Member  |
Joined: 11-Mar-2003 Posts: 76
From: UK | | |
|
| @OnlyMe
+1
I have had several Amiga's over the years and would like to have a play again, but the entry price is too high. I would want to use my machine online too, so it would also need a working browser. I don't know what the current level of browsers is but the latest time I had a machine it wasn't good (5-6 years ago).
I would come back and buy hardware and software to use in my leisure time but it needs to stack up.
|
|
Status: Offline |
|
|
olegil
|  |
Re: New hardware: PPC-Motherboard A1222 "Tabor" by Acube/A-eon Posted on 21-Oct-2015 7:36:00
| | [ #648 ] |
|
|
 |
Elite Member  |
Joined: 22-Aug-2003 Posts: 5900
From: Work | | |
|
| @broadblues
exactamundo. I've been advocating stripping it down to bare essentials here for a LONG time. Use what's on chip fully, rather than adding a lot of extras.
But it's not always that simple, so my current favourite solution would be T1022 + SB850. Drivers already exist for SB600, but getting those extra PCIe ports would be worth an update, I think. _________________ This weeks pet peeve: Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean. |
|
Status: Offline |
|
|
duga
|  |
Re: New hardware: PPC-Motherboard A1222 "Tabor" by Acube/A-eon Posted on 21-Oct-2015 8:13:20
| | [ #649 ] |
|
|
 |
Regular Member  |
Joined: 1-May-2012 Posts: 228
From: Unknown | | |
|
| @jaokim
Quote:
jaokim wrote: @iggy
Interesting!
Quote:
I doubt the difference between the e500v2 and the e500mc was of much significance. |
In lack of better things to do (or wait, I do have better things to do...) Anyhow, I looked at FreeScale's QoriQ pages, and found this chart: http://www.freescale.com/products/power-architecture-processors/qoriq-power-architecture-processors:QORIQ-POWER-ARCHITECTURE?cof=0&am=0
From that it seems evident that the cheapest e500mc comes at $71, where the P1022 e500 is at $31.
Looking at the chart, it seems that the P1022 is the only choice in the lower price segment, with GHz in par/slightly above the Sam460. The T1040 adds only $10 to the price, but perhaps it wasn't available at the time, plus that the cost increases with $10000 for 1000 units. Perhaps it was deemed cheaper to implement the FPU-fix-thingie for OS4, and start with the hardware?
Anyhoe. Need more popcorn. |
The price for P1022 is in the range $30.13-$57.92. As they have chosen dual core, 1200 MHz and DDR3 the price should be close to $57.92. |
|
Status: Offline |
|
|
CodeSmith
|  |
Re: New hardware: PPC-Motherboard A1222 "Tabor" by Acube/A-eon Posted on 21-Oct-2015 8:51:44
| | [ #650 ] |
|
|
 |
Elite Member  |
Joined: 8-Mar-2003 Posts: 3045
From: USA | | |
|
| @cdimauro
Quote:
cdimauro wrote:
That can only solve the problem of having duplicated code both for regular PowerPC and PowerPC/SPE, but doesn't solve the ABI problem which itix exposed. |
Sure it does. An ABI is just a convention - these parameters are passed in registers, that register holds the stack frame, this is what C++ vtables look like. There's no reason why one text segment could not have a different ABI to another one, as long as you don't mix and match. And even if you do, since SPE opcodes are emulated, the only penalty is that the code will run slower.
|
|
Status: Offline |
|
|
itix
|  |
Re: New hardware: PPC-Motherboard A1222 "Tabor" by Acube/A-eon Posted on 21-Oct-2015 9:01:03
| | [ #651 ] |
|
|
 |
Elite Member  |
Joined: 22-Dec-2004 Posts: 3398
From: Freedom world | | |
|
| @CodeSmith
The problem is SPEFP is still not compatible with the system ABI (calls you make to exec.librray etc). When passing floats (like to snprintf() kind of vararg calls) you have to deal with SPEFP/FPU mappings. _________________ Amiga Developer Amiga 500, Efika, Mac Mini and PowerBook |
|
Status: Offline |
|
|
olegil
|  |
Re: New hardware: PPC-Motherboard A1222 "Tabor" by Acube/A-eon Posted on 21-Oct-2015 9:18:42
| | [ #652 ] |
|
|
 |
Elite Member  |
Joined: 22-Aug-2003 Posts: 5900
From: Work | | |
|
| @duga
You forgot to finish your train of thought. T1022 at 1200MHz is $15 CHEAPER than P1022 at 1200MHz, so this would have REDUCED the cost of that production run by $15000. Hence the ONLY logical explanation can be "we cannot justify delaying this any longer". _________________ This weeks pet peeve: Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean. |
|
Status: Offline |
|
|
duga
|  |
Re: New hardware: PPC-Motherboard A1222 "Tabor" by Acube/A-eon Posted on 21-Oct-2015 9:50:06
| | [ #653 ] |
|
|
 |
Regular Member  |
Joined: 1-May-2012 Posts: 228
From: Unknown | | |
|
| @olegil
Quote:
olegil wrote: @duga
You forgot to finish your train of thought. T1022 at 1200MHz is $15 CHEAPER than P1022 at 1200MHz, so this would have REDUCED the cost of that production run by $15000. Hence the ONLY logical explanation can be "we cannot justify delaying this any longer". |
Except the fact that T1022 isn't included in the list Jaokim linked to. |
|
Status: Offline |
|
|
Spectre660
 |  |
Re: New hardware: PPC-Motherboard A1222 "Tabor" by Acube/A-eon Posted on 21-Oct-2015 10:08:06
| | [ #654 ] |
|
|
 |
Elite Member  |
Joined: 4-Jun-2005 Posts: 3918
From: Unknown | | |
|
| http://www.reuters.com/article/2012/10/10/idUS93602+10-Oct-2012+BW20121010
Quote:
Freescale plans to offer samples of the T1040 device in the third quarter of 2013, with other T1 and T2 devices following soon after. World-class RTOS and/or development tools support is planned from Enea, Green Hills Software, Mentor Graphics, Wind River and others. Freescale intends to offer development tools, including a Linux® BSP, a reference design board, a multiservice router reference design, security appliance reference design, CodeWarrior debug tools, configuration tools, models and more. For more information, visit freescale.com/QorIQ.
|
_________________ Sam460ex : Radeon Rx550 Single slot Video Card : SIL3112 SATA card |
|
Status: Offline |
|
|
olegil
|  |
Re: New hardware: PPC-Motherboard A1222 "Tabor" by Acube/A-eon Posted on 21-Oct-2015 11:15:43
| | [ #655 ] |
|
|
 |
Elite Member  |
Joined: 22-Aug-2003 Posts: 5900
From: Work | | |
|
| @duga
Yes it is. It's a variant of T1042, listed with minimum price $43@1k. Third entry from the bottom. If you limit the view to only two cores and speed to 1200MHz and expand both T1042 and P1022 you'll see them. And here I made a mistake, it's $10 cheaper, not $15.
P1022NSN2MHB vs T1022NSE7MQB _________________ This weeks pet peeve: Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean. |
|
Status: Offline |
|
|
olegil
|  |
Re: New hardware: PPC-Motherboard A1222 "Tabor" by Acube/A-eon Posted on 21-Oct-2015 11:16:45
| | [ #656 ] |
|
|
 |
Elite Member  |
Joined: 22-Aug-2003 Posts: 5900
From: Work | | |
|
| @Spectre660
Any chance of a context as to why we're reading a quote from 2012?  _________________ This weeks pet peeve: Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean. |
|
Status: Offline |
|
|
duga
|  |
Re: New hardware: PPC-Motherboard A1222 "Tabor" by Acube/A-eon Posted on 21-Oct-2015 11:28:52
| | [ #657 ] |
|
|
 |
Regular Member  |
Joined: 1-May-2012 Posts: 228
From: Unknown | | |
|
| |
Status: Offline |
|
|
Spectre660
 |  |
Re: New hardware: PPC-Motherboard A1222 "Tabor" by Acube/A-eon Posted on 21-Oct-2015 11:35:07
| | [ #658 ] |
|
|
 |
Elite Member  |
Joined: 4-Jun-2005 Posts: 3918
From: Unknown | | |
|
| @olegil
Trying to establish when the T1022 would have been available from .
Quote:
olegil wrote: @Spectre660
Any chance of a context as to why we're reading a quote from 2012?  |
_________________ Sam460ex : Radeon Rx550 Single slot Video Card : SIL3112 SATA card |
|
Status: Offline |
|
|
Deniil715
 |  |
Re: New hardware: PPC-Motherboard A1222 "Tabor" by Acube/A-eon Posted on 21-Oct-2015 11:54:42
| | [ #659 ] |
|
|
 |
Elite Member  |
Joined: 14-May-2003 Posts: 4238
From: Sweden | | |
|
| @Flasheart
Quote:
I have had several Amiga's over the years and would like to have a play again, but the entry price is too high. I would want to use my machine online too, so it would also need a working browser. I don't know what the current level of browsers is but the latest time I had a machine it wasn't good (5-6 years ago). |
The current browser (Odyssey) is good. Works with facebook, google maps, youtube, paypal, pretty much all online stores I've come a cross, online banking (if it doesn't use stinking java applets).
Few limitations, especially since Flash is on its way out from the web in favor of HTML5 Very few places use java, typically a few banks._________________ - Don't get fooled by my avatar, I'm not like that (anymore, mostly... maybe only sometimes)  > Amiga Classic and OS4 developer for OnyxSoft. |
|
Status: Offline |
|
|
iggy
|  |
Re: New hardware: PPC-Motherboard A1222 "Tabor" by Acube/A-eon Posted on 21-Oct-2015 13:43:25
| | [ #660 ] |
|
|
 |
Super Member  |
Joined: 20-Oct-2010 Posts: 1175
From: Bear, Delaware USA | | |
|
| @wawa
Soc integration helps eliminate the additional chips and added board complexity that results when features are implemented discreetly.
Its given us things like built-in NICs, USB, Serial and Parallel ports, etc.
The features aren't added 'blindly', but rather are chosen as being those most commonly needed. |
|
Status: Offline |
|
|