@gregthecanuck

I can't get any info on manufacturing costs until I can get an evaluation board at the ealiest. I wont know what chips, connectors and other components are on it until then and so won't be able to get ana ccurate quote from anyone. As soon as I know I will post it on the opfficial website as this is the cost that end users will pay (The whole idea of this project being to sell a board to users at cost whle bulk orders are sold with the normal margin which will help pay for future manufacturing costs).

U-Boot is open source and is the "universal bootloader" and so shouldn't be dificult to port. The AOS donlge code and drives will be provided by Hyperion as they are eger to spread to as many boards as possible (Please don't quote me on this, let them speak for themselves but this is the impresion I get). I have been speaking to Thomas Freiden about this project as he saw the original post and said that he was interested and would act as a mediator between me and Hyperion, I won't say any more than this as I don't want to post anything which he does not want to be said publicly, I will let him or some one else from Hyperion post on their behalf.

@Null

It would be dificult to do a cheaper G4 motherboard than G5 if the reports that the G5 CPU's are cheaper than the G4's is ture, which perosnally I can beleive given teh quantity and speed which IBM can produce them comapred to Motorola when operating as full capacity. Also being a large company would help as you can charge smaller margins.
Also, if the reports are to be beleived, IBM are trying to push the G5 mainstream by supporting motherboard manufacturers. This is a perfect opportunity to get the A1 to the forfront fo not just PowerPC technology but copmuting technology in general. The G5 is a new chip, one that is set to replace the G4 and, like the G4 did to a certain extent with the G3, sideline it to low end systems. The G5 is not just faster in terms of clock cycles and able to process larger ammounts of data because it's 64bit but it is also more effiecient per clock cycle, it has more instructuion units IIRC as well and far more advanced branch prediction logic. The G5 is a processor built to give real world performance, not be benchmark friendly and as such is blisteringly fast. The high speed bus's that are arround the 1ghz level are a huge boost over the 100mhz bus of the A1 of the 167mhz bus of the last G4 PowerMac's. The bus is not shared, each processor has iot's own for communicating with the northbridge which removes a large bottleneck from SMP systems ready for when AOS4 supports it. The Hypertransport iunterface on the Northbridge allows new southbridges to be connected, not the ancient Via ones, so old they don't even feature on the website anymore, which were connected via PCI and so hoging the bandwith that exopansion cards need. The new northbridge also features 2xPCI-X bus's for high speed 64bit PCI-X ports and an 8x AGP port, a lot faster than the 2x on Mai's Articia S or the 0x on Marvell's ;).
Also, the G5 is cheap enough now for apple to use it in their iMac's so it must be possible to build sub £1000 systems and therefore will bring the cost of A1 motherboards down (hopefully).

Seems that this community is just immune to learning ....


A few thoughts:

Buying G5s, IBM-NBs, AMD-SBs and so on in small numbers (~1000) will be rather expensive, and who would pre-finace such a prodution run (we are talking about something ~300000$) ? Doing it in even smaller numbers will again increase the price per board even further.

And that assuming that IBM/AMD will even talk to you....

You just can't develop that kinda HW from your bedroom, you'll need lots of eququiment, experience and spare parts to replace those you burned in failed attempts.

Look at what time/effort Jeri/Jens put into the the C-One, a design based on really slow clockspeeds.

Look at how "far" Oli has gotten with his CF-upgrade, CPUs that are "all-in-1-designs", where one doesn't have to hassle with NB, SB and suchlike.

Look at how much time and money bPlan invested into the Pegasos1/2, based on technology 1 or 2 generations behind the G5, and with people who have some real experience in PPC-HW-design ....

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

@Kronos

There is no reason why IBM or AMD would not communicate with us. Cna you think of one? IBM is looking to get the G5 into as many markets as possible, including a push into mainstream desktops. That means as many motherboards as possible.
Also, you don't seem to ahve a very good grasp of the way in which these companies opporate. IBM and AMD do not sell direct to manufacturers, they go through their own authroised distributers, for example, in the UK IBM's has both ACAL (www.acalsemis.com) and WBC (wbc-europe.com). There is no talking to the comapny required, you can buy their chips from distributers just like you can buy a PIC from Rapid, the only limits are the quantities, usually arround 1,000 but it depends on the company. You seem to forget that they are businesses, they are here to make money in the same way resteraunts are, you don't stop someone coming into your resteraunt just because you don't liek the look of them, if you did that you wouldn't make any money and you would get a bad reputation.
If you had read my first post you would realise that I don't just work from my bedroom and that I do have PPC expereince. Also, you do not need a lot of equipment, both maufacturing and assembly are always done by 3rd partys, even Appl don't have their own circuit board manufacturing plants, they use ones that are run by 3rd partys for the computer industry.
One of the points I would make about your examples are that they are working from scratch as such, their is no reference board for a C-One and the ones for the parts which they are using are almost certainly designed for different applications and so are litle or no use. We would be working from a reference board produced by an expereinced company who have designed many PPC and SPARC boards.

@digitaldisaster
Dunno bout your experince, but I did notice that you ask for financial help to buy a 6000$ mobo, when the whole project might burn x100 more money ....

It's like promising a new super-sports-car and than asking potential costumers for money to buy 1 ton of steel for the prototype .....

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

@Kronos

Do you think that I can afford to front money to pay for manufacturing costs, assembly costs, the required parts and the prototype. Also do you think I can relaese that money quickly and easily enough to buy a $6000 board off the bat. As I have said before I want this to be a community project. That doesn't mean the community paying for everything though. I will try to raise money from wherever I can to pay for this project, even if it means a substantial loan, but that's not something I can go and do until I know how much money I need.

And anyone that uses 1ton of steel on a sports car shoudl be shot

@digitaldisaster
Who would give someone a loan over x00000$ when said person doesn`t even have 6000$ of his own capital ? Noone ....

It might be that you have the skills to the job, and it might be that you honestly believe you will pull it through, but it is much to big for being "community-financed".

If we were talking about getting an old accel back into production, or an internal GFX-card for the A1200, than I would atleast see some reason to believe, but not this way.....

What you need:
A way to market these boards outside "Amiga", and someone believing in you enough to make a big investment on it.

Anything less is doomed to fail, and will only result in another severe hit to the Amiga-user-base, just like the BoXeR, AmiJOE, Brainstormer(Escena-A1), PIOS-one, A-Box, Pre-Box ..... and most of these didn't even include a prepay-scheme.

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

@Kronos

For once I agree with you, the Amiga/MOS market are certainly not big enough to fund such a project
--------------------------------------------------------------------------------
Just some thoughts on the G5 project.

1. I generally like the idea, very much:)
2. Getting it produced in quantities big enough to keep the price down is gonna need a LOT of money.
3. It has to be cheaper than a (well supported) Mac, if it is to be sold to these markets (Linux servers etc.).
4. copyright on the Eval, board, licensing costs?
5. raising money for a non-profit project will be next to imposibble?
6. Getting Linux people interested in the project
7. How many Amiga users are actually interested at this point (and MOS users..).
8. Are Eyetech/Kmos interested in a non-profit hardware market, will they eventually participate in the project?

What does the board cost from Momentum if ordered in larger quantities (talking 1000's instead of just one)?

Ok, this was just some thoughts, hope you get some kind of green light from Hyperion regarding OS4 support, I will certainly follow this project closely.

Troels

_________________

@Kronos

I wish I knew where you are getting these figures from, I don't know the costs at the moment, I have made enquires but I can't tell you anything until I know myself. Please be patient and until we ahve some hard numbers stop scaremongering with unjustified 5 figure numbers.
I am fully aware that to succeed it would need to be market to linux, BSD etc. customers, much as the Eyetech A1 is but I also have no real means ofdoing this and I don't think that the Linux/BSD communitites would be anywhere near as supportive of such an effort as the Amiga community, they would be even more skeptical than you are

@Troels

1)Good
2) It certainly is
3)Not nessaceraly, for some the fact that you can build your own system will be enough, as would AmigaOS support
4)The eval board is a one of cost of $6000 and has no licensing restrictions
5)Why? The kind of organisation I am talking abour is an actaul registered company but one that is set up in such a way that it may make no profits for it's members and may have no shareholders, all money musst be reinvested in the company, in effect a charity.
6)Almost certainly essential
7)I have no idea, maybe AmigaWorld.net could set up a poll. Although it would be possible to get MorphOS on it, with Genesi apparently working on their own G5 board it may be dificult to get them to support this one, especially with a different chipset. If it is true that it is the developers not Genesi/Thendic that own MorphOS then it may be possible to bypass Genesi and talk directly to the developers who I suspect would be more willing to help and less prone to empty promises.
8)Eyetech don't really come into it beyond a possible distributer unless they want to licesne the board and rebadge it as an Eyetech AmigaOne. KMOS also have little bearing on this beyond getting an AmigaOne license, unless they want to inject some funding into the project. I would contact them if my damn email server wasn't playing up.

Once again I shal try to celar up the "not for profit" bit, this is to get the const to end users (i.e. the community) down, people that make bulk orders, such as companies that would want Linux on it, would have to pay the normal markup, this money would be reinvested into the project to pay for development of new boards as well as Manufacutring costs of future batches.

I hope that their is interest in this projectotherwise I wil stop wasting my time and making a fool of myself now. I have seen a lot of posts about on Amiga forums and mailing lists say taht they want faster AmigaOne's and/or G5 AmigaOne's and would pay to help them be developed or would certainly buy one. Well if these people are prepared to put theyre money where there mouth is then this project will live otherwise I leave now and it will die. Please do not post negative comments here, this is for discussion about getting the board working and what features etc. it should have.

@Null

Quote:

Null wrote: @digitaldisaster A much lower priced G4+CPU combination would be much better than a G5 mobo. We need cheaper A1s, that's my opinion.

I think that part of this picture just isn't cheaper price. The reason is you're talking, at least in USA from Software Hut, $999 for a 933Mhz G4 w/ motherboard. Add the rest of the stuff to build a system and you're roughly at $1,500. For $500 more you could get a 1.8Ghz G5 Mac system with a better video card, Firewire 800, USB 2.0. For an additional $500, $2,500, you get a dual 2Ghz G5 system w/ PCI-X, double the harddrive, and double the memory.

Thus, while I agree that a less expensive G4 might work, if someone can keep the G5 pricing about the same while greatly increasing the speed of the system and updating the components to the newer faster standards I see this as an attractive machine. Having Amiga run a bit closer to the pack for the same amount of money is something that'll help attract power users and new users as the value/$ will be more akin to the Mac, Windows, and Linux systems available.

As for a cheaper G4 type of machine, Eyetech has said that while someone can use the Mini-ITX Amiga for their desktop machine it's not their primary market for this form factor. So someone can use the Mini-ITX for a lower cost hardware. Having a G5 ATX Amiga wouldn't step on this part of Eyetech's market. It may have impact on their G4 ATX, however. Amiga, Inc. in the past said they welcome multiple hardware vendors to create Amiga machines, so they themselves are looking for some competition within the Amiga hardware space.

Quote:

Before starting donating money to you digital, i'd wait for an official word from Hyperion.

Amiga, Inc. is the company that owns and sells the AmigaOS. KMOS owns Amiga, Inc. Hyperion is a contracted developer. I'd think Amiga, Inc.'s commitment to back the G5 would be great. I'd doubt that at this juncture Amiga, Inc. would give full commitment, but I could be wrong. However, my take is that Amiga, Inc. would need to be approached early on and give a knod to looking at the G5 solution once running to make a determination.

Another criticizm I've seen leveled here is, if you don't have enough to spend on a test motherboard you don't have enough money to spend on manufacturing the board itself. While I understand this I think there's a couple of ways to get around this point. Amiga, Inc's approval is very important first. Second how to manufacture comes into play. If Amiga, Inc. approves I'd hope they'd be willing to partner with the G5 hardware developer to find a manufacturing parter, collect some inital funding, and heck it still doesn't exclude Eyetech. Perhaps Eyetech would be willing to manufacture the product and sell through their channels. There's quite a few options here and all are important to consider. But, I believe that a working prototype is most important to be able to show Amiga, Inc and anyone else that yes it works and then all can discuss what/how this is important to the Amiga market.

@Kronos

Quote:

Look at how much time and money bPlan invested into the Pegasos1/2, based on technology 1 or 2 generations behind the G5, and with people who have some real experience in PPC-HW-design

Was Pegy II a straight ?copy & paste?(TM) engineering job from Marvell Discovery based reference board? (rhetorical question)

_________________

@BrianK

Quote:

BrianK wrote: Amiga, Inc. is the company that owns and sells the AmigaOS. KMOS owns Amiga, Inc. Hyperion is a contracted developer. I'd think Amiga, Inc.'s commitment to back the G5 would be great. I'd doubt that at this juncture Amiga, Inc. would give full commitment, but I could be wrong. However, my take is that Amiga, Inc. would need to be approached early on and give a knod to looking at the G5 solution once running to make a determination.

That is incorrect.

Amiga Inc. do not own and sell AmigaOS.

KMOS owns the AmigaOS IP.

Hyperion owns OS4 sourcecode which KMOS (originally Amiga inc.) has the option to buy back after completion.

KMOS has licencing and contractual agreements with Hyperion and Eyetech regarding AmigaOne/AmigaOS 4.x which were transfered from Amiga Inc.

KMOS owns Amiga Inc.

DigitalDisaster would need to talk to Hyperion and KMOS. Amiga Inc. wouldn't factor into it.

@Hammer

Good question ......

I think copy+paste+merge from Peg1 and Marvell-ref-board would give a realistic picture .....

And still it took ~1/2 year.

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

@Uncharted

Needless to say KMOS, Amiga Inc. Hyperion, and XYZ corps would be brought in as needed so each knows of the project and gives the necessary approvals.

@Uncharted

This is correct but I get the impression that KMOS is opperating all Amiga related activities through Amiga Inc . Whatever, I have now crucified my fachist emial server and have sent preliminary messages to both Amiga Inc. and KMOS. Once there is an AmigaOs port on a G5 eval board then we can go to companies such as IBM and ask them for financial backing to help cover initial manufacturing costs. I will contact them before hand to discuss the project with the however, I just have to find the address of my IBM contact now, off to my address book backup.......

@Kronos

This is just a guess but I suspect the Via southridge and acompnging wiring was brought over from the Peg 1 along with the CPU card edge connector pinout (but no or very little wiring) and the rest was taken from marvells reference docs apart from the AGP slot which is a FPGA which interfaces between the PCI-X and AGP slots doing the nessacery conversions

@Kronos
Quote:

I think copy+paste+merge from Peg1 and Marvell-ref-board would give a realistic picture .....

What about a straight ?copy and paste? from Momentum's PPC 970 eval board?

Note that Genesi has build AGP bridging circuits, thus it's not a straight "copy & paste" engineering job.

I defined ?copy & paste? engineering job in the level of Leadtek?s nForce2 motherboard vs NV's reference motherboard i.e. NV has stated that they are the ?closest?.

_________________

@Kronos

Quote:

Kronos wrote: @digitaldisaster Who would give someone a loan over x00000$ when said person doesn`t even have 6000$ of his own capital ? Noone ....

It's interesting you mention this, as previously you made the comparison

Quote:

Look at how much time and money bPlan invested into the Pegasos1/2, based on technology 1 or 2 generations behind the G5, and with people who have some real experience in PPC-HW-design ....

You think bPlan were funded entirely from thier own pockets? Considering their past, would you think that they'd be any more financially attractive for investors than this project? "Yeah I know what I'm doing in this field, I used to be high up in a small company that did exactly the same kind of stuff, but that went bankrupt while working on similar projects" Real inviting that is

@BrianK

Quote:

Another criticizm I've seen leveled here is, if you don't have enough to spend on a test motherboard you don't have enough money to spend on manufacturing the board itself

That particular purpose could be use for gauging real interest and spreading the risk across the community. Atm, a competitive PPC clone hardware is at its infancy compared to X86 laptop/desktop/server world.

_________________