@feanor

Welcome my friend. It's a nice addition to see people like you having visions (and greeks too!). I wish you good luck with your project. I said good luck because in the Amiga fields, you'll surely need it.

_________________

@Trixie

Quote:

If ACube actually made it in the industrial market, they would surely drop some news about that, or at least mentioned a partner. So far they haven't, so I guess no big deal from industrial sales for ACube.

Take a look at this page:

http://www.acube-systems.biz/index.php?page=company

They are mentioning some partners: they provide STx SAM motherboards which are then reselled with custom or standard configurations, no mention if they are intended for industrial market or not. They also mention Denx, but not sure which way the collaboration goes between them and ACube. Also, it is a fact that they spend only very little time for PR management, likely because they're a small company. And personally, in general when I don't have much information I tend not to make any assumptions.

@Crumb
Quote:

Since industrial customers don't have any advantage using expensive custom PowerPC boards I doubt it's sold more than in the OS4 market. Otherwise Sam440 price would fall (and it's as expensive as in the first day)

I'm not sure that you can only factor price to be the only reason for the motherboard to be attractive. What about motherboard configurations (FPGA, possibility to mount ccTalk, LVDS and SmartMedia connectors), customers already running PPC solutions and not willing to do a CPU switch, local production (they are produced in Italy, low waiting times/better support might be a factor?), available quantities (they might be willing to produce very small batches), and there might be even more specific factors that I can't think of right now.

Regarding the price drop, isn't the non-flex Sam440 out of production now? If so, it wouldn't have had time to get a price reduction. Another consideration is that the final price might be nearly completely dictated by the components' companies, which may not make any discount even if the component isn't anymore cutting-edge (and it might even be the opposite).

Varthall

_________________
AmigaOne XE - AmigaOS 4.1 - Freescale 7457 1GHz - 1GB ram

@Zardoz

4º time and you search whatever excuse to not recognize it. But facts are stronger than them:

Quote:

Vision: E-SATA CONNECTIONS (which include also USB):

Quote:

Zardoz: eSATA with USB? Never seen that anywhere and that Wiki does not claim it either

Proofs of the first sentence:

http://laptops.toshiba.com/innovation-lab/esata-usb-port

http://www.hitechreview.com/uploads/2009/05/msi-power-esata-11.jpg

http://www.ohgizmo.com/2009/05/14/msi-unveils-power-esata/

I will repeat it for last time: There are eSATA connectors which include USB. You can connect any sata device WITH A STANDARD CABLE, or you can connect any USB device WITH A STANDARD CABLE. If someone want to make a new motherboard (even Amiga one), can include this technology easily. If someone thinks this didn´t exist: HE WAS WRONG. The discussion you started against me was about if those connections existed or not, not if they are standard in the straightest sense of the word, which is not relevant in any case.

Do they exist o not? they do. I don´t care what you will keep saying from now on.

I understand it is hard when someone corrects you, but there is also a very old saying which says "rectifying is for wise people"


@_Steve_

I know already that in this web, discussing against a moderator is a lost battle even before starting. And not only me, but also (sadly) a lot of people. Zardoz was the one who started a discussion denying my word about something which proved to be real, not me. And keeps insiting and editing his posts. Anyway, helping a friend is understandable, denying basic facts, not.

@vision

Quote:

vision wrote: @_Steve_ I know already that in this web, discussing against a moderator is a lost battle even before starting. And not only me, but also (sadly) a lot of people. Zardoz was the one who started a discussion denying my word about something which proved to be real, not me. And keeps insiting and editing his posts. Anyway, helping a friend is understandable, denying basic facts, not.

_Steve_ was refering to your tone. To a native English speaker post #37 was a bit snarky. Perhaps his spelling out dictionary terms in post #28 set you off in that direction a bit. Whatever. Everyone just play nice now ok?

_________________
Pegasos2 G3 running AOS 4.1 and MorphOS 2.0
Amikit user, tinkering with Icaros VM (AROS)
EFIKA owner
Amiga 1200

@feanor

Quote:

ut I will definitely fund the Haiku porting, because I like it and I think it has a lot of future.

Do you have any idea how quickly a Haiku port would be available after the hardware was shipping to customers? I ask because the x86 alpha took years to come about. Have you approached them on this yet?

Is there any other OS that it will support (like a flavor of Linux) that it will ship with?

_________________
Pegasos2 G3 running AOS 4.1 and MorphOS 2.0
Amikit user, tinkering with Icaros VM (AROS)
EFIKA owner
Amiga 1200

@many

Ok, I appreciate the noise around this, but things are pretty simple. The system is not going to be a PC-killer (I never intended to do that, it would be folly without huge pockets to do it). Whatever the CPU used it will contain things that are now *standard* in the PC/Mac world (well, now with the release of the first USB 3.0 controller chips) it might *just* be possible to manage to squeeze one in the specs. If the companies (NEC for one) even consider to speak with us.

In any case, this is not vapourware, and whether it happens or not it will depend on the user feedback I get. So far, response is good -I'll blog about it in the weekend, with actual numbers/stats- but it barely is enough to warrant a market. But it's very early to tell, and I didn't expect much more to be sincere.

Anyway, let's wait and see. The board will most likely come to life one way or the other. I'm determined to make one. The users however will influence greatly which way it will go. If I get tons of Amiga/MorphOS mails asking for AltiVec, then I'll probably choose one of the MPC86xx CPUs. If however I get mostly mails asking for low price/power/size then the P1022 would be a more probable choice. There are other factors, still, but you get the idea.

Konstantinos

@Varthall

Having a PPC and short supply is usually not a good strategy to sell to industrial customers... what happens if the hardware breaks? how's the client supposed to exchange his Sam440?

There are lots of companies selling "custom" boards with fpgas and exotic connectors.

Anyway I think you can put an off the shelf PC with some Altera on PCI card and obtain better performance, easier replacement if something goes wrong and better price.

@feanor

Back to topic...
A new board should offer at least the same performance as the 2nd hand options and a reasonable price. Last time Genesi planned to release a QuadCore G5 it was going to cost around 999$. I think that producing a PPC board with a single 2Ghz 970MP cpu may be reasonable if bPlan is behind the design. Avoid producing various models because it makes the platform more expensive and avoid producing boards with cpu slot as that decreases reliability and increases (design and manofacturing) costs.

IMHO a 1/1.3Ghz board is not interesting for the market: we'll have 1.5Ghz Mac Mini in less than 3 months.

BTW, Altivec is a must if we want the machine to play 1920x1080 H264 videos.

Last edited by Crumb on 24-Sep-2009 at 11:25 PM.
Last edited by Crumb on 24-Sep-2009 at 11:24 PM.

_________________
The only spanish amiga news web page/club: CUAZ

@fairlanefastback

Quote:

fairlanefastback wrote: Do you have any idea how quickly a Haiku port would be available after the hardware was shipping to customers? I ask because the x86 alpha took years to come about. Have you approached them on this yet? Is there any other OS that it will support (like a flavor of Linux) that it will ship with?

There is already a powerpc port in the works, all I have to do is fund the port to give priority to the -future- board. Some haiku devs probably know about this already as I posted a similar inquiry to the haiku forums -though being mostly x86 i was pretty much deterred from my venture :).

In any case, I don't expect it to be instantly available, but by the time first mass production run finishes it should be mostly workable -one of the haiku devs should get a prototype.

Linux support is without question. I've been a Linux developer for ages, in fact I was an official Debian Developer 1999-2008, so building a Debian-based distro or adding support on the official tree should not be that hard. I also have good relations with the Gentoo/PPC and the OpenSuse/PPC folks. So there. :)

Konstantinos

Last edited by feanor on 24-Sep-2009 at 11:36 PM.

@feanor

Your avatar should be Prometheus,the crows are already going for your liver!

I am afraid that video less that full 1080 HDTV will doom any NEW board,of course I would expect video to be done with a standard? card.

@vision

Quote:

I will repeat it for last time: There are eSATA connectors which include USB. You can connect any sata device WITH A STANDARD CABLE, or you can connect any USB device WITH A STANDARD CABLE. If someone want to make a new motherboard (even Amiga one), can include this technology easily. If someone thinks this didn´t exist: HE WAS WRONG. The discussion you started against me was about if those connections existed or not, not if they are standard in the straightest sense of the word, which is not relevant in any case.

Let's reread that quote of mine you pasted:

Quote:

eSATA with USB? Never seen that anywhere and that Wiki does not claim it either

Underline "seen". Hadn't seen them. And the wiki you liked to indeed does not claim such a thing as a standard either.

For the n-th time, I think there is a language barrier here. Yes, I hadn't seen those before. Yes, they do exist, as you very obviously and correctly pointed out.

HOWEVER:

Quoting YOU:

Quote:

They do baby: http://laptops.toshiba.com/innovation-lab/esata-usb-port . And of course not just toshiba, but all eSATA controllers since the first revision.

Wrong. They do not. eSATA controllers do not include USB, Toshiba, MSI and others have made a connector that is compatible with both USB and eSATA. They also make some custom hardware and cables that connect to both at the same time. Of course you can connect to a SATA device with a standard cable, from the point of view of the female eSATA connector it's just an eSATA port and from the point of view of a standard female USB connector it's a standard USB connector. However, neither the USB nor the eSATA connectors "know" anything about each other, it's a standard eSATA controller and a standard USB controller, wired to one combo port that can accept either cable plus some custom (non-Standard) cables that draw power from the USB port. Powered eSATA is not a recognised standard by the committee that makes the eSATA standards, it's an MSI thing.

Which, of course, is exactly what my reply to that stated:

Quote:

What you are talking about is nothing more than a combo port, USB on one side, eSATA on the other. It has the advantage of providing power and saving space, although that'd only be useful for 2.5" SATA drives, as 3.5" require 12V and significantly more current than USB can provide. It's this connector.

Next one:

Quote:

The connector is THE SAME (where did you get your "sides" from? wuahahaha- your are really funny-) and I know it because I am posting this from a computer which delivers it from factory and I use it sometimes as eSATA and sometimes as USB, and it is from the "medion" brand (no cutting edge technology precisely). Are you trying to deny the existence of what I am already using every day???? I repeat to you again (hope this time you get it): THIS IS A STANDARD CONNECTOR. It is the first revision of the eSATA protocol, called (as your own link shows, thank you!) POWER ESATA. Is it clear now?

Again, you were wrong. Which we established, of course, and in fact you admitted in this post. It is not a standard connector. It's an MSI connector. The standard power connection for eSATA does not exist yet, its development has been announced by the relevant parties. Now, I think you'll agree with me that the SerialATA working group would most definitely not announce a Power over eSATA standard if they already had one.

Quote:

I understand it is hard when someone corrects you, but there is also a very old saying which says "rectifying is for wise people"

I have no trouble admitting to be wrong, as I have done in this thread already, I did not know USB combo ports existed, you pointed them out, fine. I also did not know that eSATA had different signal levels, as I'd only ever used passive adapters. I was however 100% correct that your combo connector is not a recognised standard connector and I was correct that USB is not part of eSATA controllers. _Steve_ got you more links reinforcing those. Or do you seriously think that he agreed with me because "I'm his friend" and won't admit I'm wrong? So far, I have admitted to be wrong twice in this thread, and you haven't done it once.

Quote:

I know already that in this web, discussing against a moderator is a lost battle even before starting. And not only me, but also (sadly) a lot of people. Zardoz was the one who started a discussion denying my word about something which proved to be real, not me. And keeps insiting and editing his posts. Anyway, helping a friend is understandable, denying basic facts, not.

Ok, you really have issues with the English language. You started this with this:

Quote:

Review your technlogy knowledge (MUA-HA-HA)

That is extremely snarky and, OK, I admit, that made me lose my cool. But I did not moderate you and I would not moderate you, as I was involved in the discussion and directly fed it. You continued with very similar posts.

BTW, any editing I did not remove information. I'd written a post that only answered one part of your message and then replied to the rest. The other edit is about a technical mistake in signal levels and is clearly marked with Edit:.

EDIT: Just so you don't complain that I edited my post: I corrected "posts" in the penultimate paragraph, I'd spelt it as "ports".

Last edited by Zardoz on 24-Sep-2009 at 11:42 PM.

_________________

@Crumb

Quote:

Crumb wrote: @Varthall Having a PPC and short supply is usually not a good strategy to sell to industrial customers... what happens if the hardware breaks? how's the client supposed to exchange his Sam440?

I suppose with a newer, but compatible motherboard, such as a flex one.

Quote:

There are lots of companies selling "custom" boards with fpgas and exotic connectors.

And can they cover *all* the needs of industrial customers?

Quote:

Anyway I think you can put an off the shelf PC with some Altera on PCI card and obtain better performance, easier replacement if something goes wrong and better price.

Again, are the price and performance really the only factors that make a motherboard attractive in that market?

Varthall

_________________
AmigaOne XE - AmigaOS 4.1 - Freescale 7457 1GHz - 1GB ram

@BigC

Quote:

I am afraid that video less that full 1080 HDTV will doom any NEW board,of course I would expect video to be done with a standard? card.

Well, it has been demonstrated elsewhere (youtube iirc) that the 8610 could play 1080 HD videos, so I guess it's possible. I'm pretty sure the 8640D would do that without sweating as well. But I'm not so sure about the P1022 since it has no AltiVec unit to help (the SPE is ok, but not that powerful). But as you say, a gfx card would help there, and whatever the choice of CPU, the board will include a gfx PCI-e connector (x16 iirc, but with the largest available lane for each CPU, x8 max).

Konstantinos

@feanor

Apologies about the mess earlier.

Anyway, it is an interesting effort. And the D at least shouldn't really break a sweat indeed. It's be a useful little platform for playing with the PowerPC architecture at a decent frequency with Altivec. As you've correctly identified, though, you'd have to find more than just one target market that will actually pay for it even in smallish numbers.

Alkis

_________________

@feanor

I'll add my 2c...

I think it's a good thing what you're trying to do, and providing hardware for more than one hobbyist group is a smart decision. I wish you the best of luck, for all our sakes.

Assorted stuff:

1. People mentioned putting the PCIe x8 in a x16 slot, and leaving a gap between that slot and the next one, and I agree that those are very good ideas.
2. Put as much "standard" functionality on the board as possible - ideally, one should be able to run the board with just a video card plugged in and not miss anything. That will cut down on the number of device drivers that need to be written for the board to be usable. The number of people who can write professional grade AmigaOS device drivers can probably be counted on the fingers of one hand.
3. By that logic, if you can put a decent video chip on the board itself (eg a mobile Radeon that doesn't need a huge heatsink+fan), so much the better.
4. AmigaOS does not currently support multicore, and some knowledgeable people claim that it would be impossible to do so without breaking backward compatibility, so at least for the first revision concentrate on single core.
5. People mentioned talking to ACube, Genesi and Hyperion. Those three are the established big players in the post-CBM Amiga world (Genesi not so much any more, but it was recent enough that the experience is probably still in the company).
6. I don't know if you've talked to them already, but stop over by morphzone.org and aros-exec.org - those are the equivalents of this site for morphOS and AROS, two OSs that are compatible with AmigaOS, which arose from the post-CBM ashes. I bet you'll find supporters for your project there as well.

@CodeSmith

Quote:

1. People mentioned putting the PCIe x8 in a x16 slot, and leaving a gap between that slot and the next one, and I agree that those are very good ideas.

Actually, some boards use a slot with the terminating plastic missing. That allows putting a 16x card even in an 1x slot and provided there are no obstructions on the PCB, such as heatsinks, that could save quite a lot of PCB real estate.

_________________

@CodeSmith

Quote:

CodeSmith wrote: @feanor I'll add my 2c... I think it's a good thing what you're trying to do, and providing hardware for more than one hobbyist group is a smart decision. I wish you the best of luck, for all our sakes.

Thanks :)

Quote:

Assorted stuff: 1. People mentioned putting the PCIe x8 in a x16 slot, and leaving a gap between that slot and the next one, and I agree that those are very good ideas. 2. Put as much "standard" functionality on the board as possible - ideally, one should be able to run the board with just a video card plugged in and not miss anything. That will cut down on the number of device drivers that need to be written for the board to be usable. The number of people who can write professional grade AmigaOS device drivers can probably be counted on the fingers of one hand.

totally agreed and noted.

Quote:

3. By that logic, if you can put a decent video chip on the board itself (eg a mobile Radeon that doesn't need a huge heatsink+fan), so much the better.

I think that's a bad idea. The lowest one can get is a Radeon HD 2400 (iirc), which is ok, but not a speed daemon. It should work for most people, but it would also raise the cost for the ones that will plug a newer and faster gfx card anyway. For builtin display I think you can't beat the MPC8610, which unfortunately doesn't have a 3D unit.

Quote:

4. AmigaOS does not currently support multicore, and some knowledgeable people claim that it would be impossible to do so without breaking backward compatibility, so at least for the first revision concentrate on single core.

I think that should be decided after consulting with Hyperion, which I will do in the next days.

Quote:

5. People mentioned talking to ACube, Genesi and Hyperion. Those three are the established big players in the post-CBM Amiga world (Genesi not so much any more, but it was recent enough that the experience is probably still in the company).

I'm talking with Genesi folks pretty much every day and I have cooperated with them in the past as well -and probably will so in the future as well, the board will be made by bPlan. :)
I admit I haven't talked to ACube and Hyperion because I didn't know what kind of response I would get in the forums. It seems I should talk to them.

Quote:

6. I don't know if you've talked to them already, but stop over by morphzone.org and aros-exec.org - those are the equivalents of this site for morphOS and AROS, two OSs that are compatible with AmigaOS, which arose from the post-CBM ashes. I bet you'll find supporters for your project there as well.

I've already asked on morphzone.org, I didn't think about aros-exec.org, I admit. I also have asked on haiku-os.org -with not that great success I admit- and I plan to also send mail on linux/powerpc forums/lists. So I have quite enough work to do already. The goal is to reach the 500-mark (even less would do, but the 500 is a min production run in a board factory).

Konstantinos

Last edited by feanor on 25-Sep-2009 at 12:08 AM.

@minator

The 476 sounds like an interesting processor. I just came across it yesterday on ppcnux.de and initially had the same idea: "Yay - an el cheapo ppc by ibm! Bring it on!" But then I realized, that it will come in Q4 2010.
Well, one should keep that cpu on the radar, but for *now* it is no option.

_________________
My programs: via.bckrs.de
MorphOS user since V0.4 (2001)

@feanor

I'm glad to hear that Genesi is back in the Amiga game! their latest announcements made it seem (to me, at least) that they were going to concentrate on ARM instead of PPC.

When you talk to Hyperion, make sure that you get them to explain (even if it's under NDA) what exactly the situation is right now between them and Amiga Inc. With the release of OS4.1 for the SAM440 (and the SAM Flex, which is bundled with AmigaOS 4) it seems like things have normalized, but there have been no official statements from either company so clearly some things are still unresolved. I'd hate to see you get burned and turned off the Amiga for good.

@CodeSmith

Quote:

CodeSmith wrote: @feanor I'm glad to hear that Genesi is back in the Amiga game! their latest announcements made it seem (to me, at least) that they were going to concentrate on ARM instead of PPC.

Ehm, I didn't say that, I said that I'm cooperating with them about the new board, it will be design/produced by bPlan but it will be sold/marketed by Codex (my company, and development will be done here as well). Just so that there is no misunderstanding here. For Genesi, AFAIK, they are focusing on ARM -and I agree with that move, actually- but I do know that they want to work with PowerPC again in the future.

Quote:

When you talk to Hyperion, make sure that you get them to explain (even if it's under NDA) what exactly the situation is right now between them and Amiga Inc. With the release of OS4.1 for the SAM440 (and the SAM Flex, which is bundled with AmigaOS 4) it seems like things have normalized, but there have been no official statements from either company so clearly some things are still unresolved. I'd hate to see you get burned and turned off the Amiga for good.

Noted. Thanks. Though I've been following the situation now and then -being an old Amiga user myself- and it's not like I'm in the complete dark :)

Konstantinos

Last edited by feanor on 25-Sep-2009 at 12:34 AM.

@feanor

Quote:

feanor wrote: Quote: 4. AmigaOS does not currently support multicore, and some knowledgeable people claim that it would be impossible to do so without breaking backward compatibility, so at least for the first revision concentrate on single core. I think that should be decided after consulting with Hyperion, which I will do in the next days.

I'd welcome a multicore machine, even if I knew that only one core would be usable for the time being. We're never going to get multicore support without multi-core hardware. Unless you can find pin-compatible single and multi-core chips, just stick the multi-core one on the board.

Hans

_________________
http://hdrlab.org.nz/ - Amiga OS 4 projects, programming articles and more. Home of the RadeonHD driver for Amiga OS 4.x project.
https://keasigmadelta.com/ - More of my work.