@iggy

But of course it depends on the codec used.

Quote:

iggy wrote: @Spectre660 Quote: Single core can playback 720p. OK...I'll keep that in mind when we get support for this board...

_________________
Sam460ex : Radeon Rx550 Single slot Video Card : SIL3112 SATA card

@Spectre660

How the file is encoded is pretty significant.
Some HD content really increases cpu load on my systems, so I understand what you are saying here.
If Tabor can handle 720p resolution video, that is pretty good.

The SAM440 isn't good at it, and the SAM460 is marginal.

As has been said here before, bottom line, its the performance that matters.

Last edited by iggy on 25-Oct-2015 at 10:18 PM.

I know that it's a risk long shot, but I guess that the only way that A-EON is gonna sell enough of this device is if they'll stop supporting 68k Amigas and giving a discount for anyone upgrading from 68k to Tabor. I'm not saying they shouldn't do it, but I see that that may be *the* solution to get some folks over the barrier of upgrading to something new.

_________________
Proud user of AmigaOS 4.1 on an AmigaONE 500. This is the first Amiga I've ever had so I don't know all the ins and outs of AmigaOS yet, so I'm sorry if I'm asking noob questions and stuff.

@Vistaus

Quote:

Vistaus wrote: I know that it's a risk long shot, but I guess that the only way that A-EON is gonna sell enough of this device is if they'll stop supporting 68k Amigas and giving a discount for anyone upgrading from 68k to Tabor. I'm not saying they shouldn't do it, but I see that that may be *the* solution to get some folks over the barrier of upgrading to something new.

that has been the policy so far anyway. and here we are now. or is there some product i missed on?

@wawa

Quote:

that has been the policy so far anyway

Since when?
I don't remember them dropping 68K support, nor have I ever been offered any discount.

@iggy

Quote:

iggy wrote: Yes the design work necessary to change the cpu on this board could have been started almost two years ago.

Provided they were actually aware of the e500v2 problems. Somehow it feels like they weren't? Which is kind of strange, given the fact that this subject was discussed at length at multiple forums back when the "Ultimate PPC" was announced. I know for a fact that Trevor saw this back then. Which is why it's such a mystery we actually see a e500v2 based Tabor today. Didn't he learn anything back then? Did he discard the discussions as "trolling", or what? It's difficult to find a plausible explanation to the Tabor situation...

@kyle

Quote:

kyle wrote: Is it so difficult to understand that when development of Tabor begun the T1xxx processor didn't exist again?

It's a very poor choice of CPU anyways, there were other CPU's at the time that were actually fully "PowerPC compatible", and several even better CPU's followed shortly after. From this thread we have learned that there is no OS4 for the Tabor board yet, and from history we know that it usually takes between 1-2 years for the OS4 people to support new H/W (provided that there still are people among the remaining OS4 developers who has this competence). So it should be possible to design a new motherboard with a proper CPU starting today, and still have it finished before OS4 is ported to it (using a development/reference board for example). The Tabor was an error, a mistake. It would be an even bigger mistake to multiply it by a thousand, as has been suggested in this thread. Just sayin' of course, it's not my money on the stake after all.

Quote:

Sure, now it could (how easy is playing with others' money) but the new Tabor probably would not be ready before 2 years...

It's a SoC, meaning it's not a lot of external controllers or components involved (relatively speaking compared to traditional computers/motherboards). You put the chip on a PCB and route traces to memory, connectors etc. Not exactly rocket science for those who knows this stuff, and there is no reason to why it should take 2 years instead of a few months. Porting the OS4 is a bigger problem and probably more time consuming. Especially so with the e500v2 core (depending on what result you aim for of course).

@umisef

Quote:

umisef wrote: Quote: why the Hyperion developers don't have a software solution for hardware that apparently exists. Because someone needs to implement such a solution. And before anyone can implement such a solution, someone needs to create a suitable kernel for OS4. And the "kernel guy" might just be busy with something else. That's what gets me about all these discussions about how well or not such a board could theoretically be made to work under OS4 --- they ignore the realities of the work involved in getting OS4 running on it at all, and the observation that even that work is apparently, uhh, "taking its time", to be generous.

Very well put!

@Spectre660

In post #788 you post a link to a video, in which that Freescale manager shows us how the P1022 plays back a H.264 video stream using the CPU (single core). He explicitly explains that the resolution is 800x600, which is (not extremely far off) half the size as a 720p stream.

In post #799 you claim that the P1022 can play back a 720p stream in S/W using the CPU (single core). ()

Then in post #801 you state the obvious: "But of course it depends on the codec used."

Well, DUH!!! Of course it does! The Efika 5k2 could decode/play back DVD resolution MPEG-2 streams. But the discussion you raised yourself in post #788 was based on the H.264 codec, and H.264 (or X.264) is what people normally use when discussing benchmarking based on video decoding/playback!

The performance in this regard is rather well established for various PPC CPU's/systems.

For example, a Mac Mini PPC @ 1.42GHz+ can play back (most) X.264 streams in true 720p without frame skipping. And this is a real G4, with proper FPU and Altivec enabled!

And from this very thread we have learned that the "Tabor" MPlayer performance in single core is comparable to a Mac Mini PPC @ 600MHz. Very far off from 1.42+ GHz! But kind of in line with what the Freescale manager in "your" movie is demoing, right?

I hope you are confused (perhaps by GPU assisted playback in your Linux setup). Because otherwise it feels like you are misleading and/or lying. And that wouldn't be very nice!

Last edited by TRIPOS on 31-Oct-2015 at 10:48 PM.

@pavlor

Quote:

pavlor wrote: MPC8610 lacks USB and SATA (not a problem with PCI/PCIe).

The way this was usually handled was by adding a southbridge.




Quote:

Quote: Back in 2006 Genesi was also planning something along the 8610 (not 100% sure about the number) as a followup to the Efika5200. They had several projects: one with G5

"The Open Server Workstation is a six layer board with two 970MP processors, the CPC945 and Broadcom's HT-1000 and 2000 chips. Excepting those parts, the board and component cost is below $200.





IIRC it eventually turned out that IBM wasn't willing to negotiate price in the way Freescale had done in the past, and the whole thing would have ended up being quite expensive. They had hoped to introduce a full system at $1,500 but when it turned out it would have become more expensive than this, they decided it was an impossible product in 2006 (not to mention the few coming years from then on).

This also made them re-evaluate the desktop/workstation market in general, and the PPC based desktop/workstation market in particular. They had already begun this process but eventually they refocused entirely on the Efika concept products (such as the 5200B, the 5121e and 5125, and later ARM based machines).


Quote:

one with 8640/8641D

They had development boards which they evaluated but IIRC they scrapped this idea in favor of the G5.


Quote:

one with 8610

Actually they considered two!

First they looked for a PPC alternative to the 5121e (after that "Lime PC" project crashed and burned). Something like this:




Much later they proposed a community funded motherboard (because they did no longer believe in a "G4" Desktop motherboard themselves). This as a response to a desire from community members who came up with the idea.

The result would have been an Open Source motherboard with:

- Flex form factor - MPC8610 @ 1Ghz (or faster if it would have turned out economically feasable)
- 4x SATA 2 connectors
- 4x USB2 ports
- 2x Gigabit Ethernet ports
- Sound ports (Audio Out, Mic In)
- 1 PCI slot
- 1 PCI-e slot 1x
- 1 PCI-e slot 8x (for GFX card)

The funding for this entire R&D project would have ended at a cost comparable to only 20 units of "X1000".

Unfortunately the community obviously didn't look at this as a good idea, so it stranded. A shame, really.




Quote:

and one with 5121e.

Actually, it was supposed to have been an entire family of consumer products (six of them), and most relevant for "us" would have been the "thin desktop" (smarttop) and a "thin laptop" (smartbook).

This was what later became "Lime PC" / "Cherry Pal" (Yes, the laptop of that product family is "the mysterious OS4 netbook that never was released" that has been discussed to death here over the years)

Here is the Lime PC "desktop" computer and its keyboard.



Note the $99 Efika 5200 also present in the picture. The purpose of the Efika 5200 was originally mainly to provide a S/W development platform until the 5121e chips/products became ready. In the end it turned out that the 5121e wasn't that compatible, and not even cache coherent.


Quote:

Then they leaved PowerPC market for ARM

And here they actually accomplished the products they tried to do with what became the "Lime PC", at a much better result!



The moral of this story:

1. Sometimes it makes sense to cancel unrealistic or impossible products

2. It was impossible to bring realistic PPC offerings to compete on the desktop market in 2006 already

3. PPC is an impossible architecture altogether for the products "we" desire, and this is even more true now in 2015, and it's a truth that becomes more and more painful annually

4. The prices some people in this community consider "fair" is in reality obscenely high to anyone else not indoctrinated on Amiga community web forums. The Genesi products (those who really were released, and those planned but later shelved) never came close to the prices that has now almost become "normal" on AW.net and A.org, and these were a decade ago!

@TRIPOS

Your long and richly illustrated post shows how important is to think beyond normal business. You can´t pretend to be "saviour of Amiga" and then leave this platform for better profits elsewhere. Genesi´s use of Amiga (MorphOS) community was clever and unscrupulous, but in the end tarnished Genesi´s reputation in eyes of many of its former supporters.

@pavlor

Quote:

pavlor wrote: @TRIPOS Your long and richly illustrated post shows how important is to think beyond normal business. You can´t pretend to be "saviour of Amiga" and then leave this platform for better profits elsewhere. Genesi´s use of Amiga (MorphOS) community was clever and unscrupulous, but in the end tarnished Genesi´s reputation in eyes of many of its former supporters.

Yes, unlike entities like Amiga Inc, Hyperion and Eyetech, whose reputation is/was "solid", right?

Genesi left PPC a long time ago, for obvious reasons (don't know if Genesi still exists today though). The MorphOS Team has been working towards an emigration away from the PPC since 2011, and they have been introducing support for more than 80 widely available mainstream PPC systems in the meantime (currently more than 3,000 OS licenses sold, pace increasing), representing the very best the PPC had ever to offer desktop/laptop consumers. No MAI Articia-S, no "Xorro/Xena", no e500v2. No €3,000 systems with a fraction of the performance compared to above mentioned second hand mainstream systems.

My "between the lines" point with post #810 was to illustrate the different paths (and the different success) of "Team Red" vs. "Team Blue".

Why is "Team Red" so reluctant to face reality and reconsider? Why is €3,000 X1000 systems and e500v2 Tabor systems "the savior of Amiga", when it's really what is killing it off?

These are the biggest unanswered questions.

Last edited by TRIPOS on 01-Nov-2015 at 01:00 AM.

@pavlor

Quote:

Genesi´s use of Amiga (MorphOS) community was clever and unscrupulous

Why?
Because a sub $600 motherboard was made available that didn't include the use of the defective Artica S?

Or was it the even cheaper design they introduced later?

They , like all companies, had to generate a profit.
So they shifted focus to ARM.
I hold no grudges.

If Hyperion had bothered to support the Pegasos II when it was still available (instead of waiting until years after), maybe a few more would have been sold.

I wouldn't attack Genesi for its practices, when the 'real' Amiga community was being supplied with substandard crap that continued to use the Artica S.

Come to think of it, what was the course that was followed by 'Amiga'?

Oh yeah, the rights were bought by a former Gateway truck driver, who contracted a games company to write the OS, which took almost ten years, and then said game company (run by a lawyer) used legal BS to screw the a fore to mentioned truck driver out of part of his intellectual property.

There's some "good" people there.

Right now, outside of A-eon (which just made its first real blunder) there doesn't seem to be anyone at the core of the Amiga push that is really trustworthy.

Last edited by iggy on 01-Nov-2015 at 12:21 AM.

@TRIPOS

Quote:

Why is "Team Red" so reluctant to face reality and reconsider? Why is €3,000 X1000 systems and e500v2 Tabor systems "the savior of Amiga", when it's really what is killing it off?

Can't agree with that either.

The X1000 was an excellent design.

Two things tainted it.
First, the design is really derived from a spec produced by Ack Systems for Amiga Inc.
Second, Ben Hermans was still involved in the older version of A-eon.

So, there was no way MorphOS developers were touching it.

But, both OS4 and MorphOS run on PPC hardware.
The SAM460 now has MorphOS support (even if that needs better graphic and sound support to be really competitive).
AND the X5000 WILL be supported.

Lest we get too partisan, none of the entities you've mentioned want to help "kill off" the Amiga.

There has been some damned nice hardware created for us to play with.

My sole objection is the CPU chosen for Tabor.
I'm not anti OS4 (except for the fact that Ben Hermans is NOT my favorite person in the world).
Nor am I anti-PPC (I just don't like the e500v2 core).

But come on!
Flawed hardware is worth pointing out!

@iggy

Quote:

The X1000 was an excellent design.

Except that it wasn't. It was a design resulting in a €3,000 system with a (single core) performance below a G4 Mac Laptop. IMHO, it should have been canceled even before prototyping. So I guess we don't share the definition of the word "excellence".

@TRIPOS

Quote:

I guess we don't share the definition of the word "excellence"

No, we don't.
It was dual core, with two slots for Graphics cards, and featured a very power efficient processor.
And unlike the processors included in the X5000, it had AltiVec instructions.

Expensive, yeah, a board with a $600 processor is likely to be that.
But the thing is, Trevor took an idea that was probably never going to get off the ground if Ainc had to be relied on, and he made it real..

And while I will miss AltiVec in the X5000, that should perform better at 2/3 the price.

Finally, if a board had been built with a T10XX processor we could have had a 64 bit dual or quad core board that could have been offered at only a few hundred.

Instead, we have Tabor (which I personally don't want no matter what the price is).

So, I might over spend on an X5000, as a last momento of our current ISA and the state of our hardware development.

Or not (after all G5 Macs ought to be a close match).

And then I'll wait for the X64 shift.

Can someone summarize this thread please.... So Tabor for Amiga OS.... is it a good idea/purchase?

[YES]

[NO]

Thanks :)

_________________

@Everblue

YES .....

Well you did ask!

TrevorD

_________________
No, I don't need no reason, I'm just breezin'

@iggy

Quote:

iggy wrote: @TRIPOS Quote: I guess we don't share the definition of the word "excellence" No, we don't. It was dual core,

Still unused.
Quote:

with two slots for Graphics cards,

Is this supported (or will be in "two more weeks")by the o.s.?
Quote:

and featured a very power efficient processor.

The X1000 target is desktop, not mobile. And, if I remember correctly, they had problems to run the processor at the maximum speed.
Quote:

And unlike the processors included in the X5000, it had AltiVec instructions.

And an accelerator unit for the FFT, but you need proper code to use it. So it's questionable how good is including this unit instead of providing a more powerful Altivec unit.
Quote:

Expensive, yeah, a board with a $600 processor is likely to be that. But the thing is, Trevor took an idea that was probably never going to get off the ground if Ainc had to be relied on, and he made it real..

Nevertheless, you should take a look at the market and build a product which fits on it. The X1000's CPU is expensive and performs on average BELOW an old G4 AND there were already G4s (even dual core) available at the time, that were less expensive.
Quote:

And while I will miss AltiVec in the X5000, that should perform better at 2/3 the price.

Let's wait for real-world benchmarks.


@Everblue

Quote:

Everblue wrote: Can someone summarize this thread please.... So Tabor for Amiga OS.... is it a good idea/purchase? [YES] [NO] Thanks :)

If you don't plan to use floating point-intensive applications, then YES. Otherwise it's a NO.

P.S. That's because the FP unit is weak, and regular FPU emulation is and will be deadly slow, no matter what Hyperion can declare about it; you'll see.

@cdimauro

Quote:

Is this supported (or will be in "two more weeks")by the o.s.?

Do you even know topic you are talking about? Two PCIe slots were blessing for further developement of RadeonHD driver. So yes, you can use two PCIe GFX cards in OS4.

Quote:

And an accelerator unit for the FFT, but you need proper code to use it.

From benchmarks we saw, no specific PA6T compiled binaries were needed.

Quote:

there were already G4s (even dual core) available at the time, that were less expensive.

PA6T has much better expansion capability than MPC8640D.

Quote:

If you don't plan to use floating point-intensive applications, then YES. Otherwise it's a NO.

As someone other in this thread wrote: Let's wait for real-world benchmarks.