@olegil

Hey this sounds really interresting. Maybe a chip for OS4 until 8641D arrives
It would actually only need a couple of support chips to be a complete desktop board.
Sound chip (I know you don't favour this) and some PATA/SATA chip.
Maybe an extra USB chip for additional USB ports, two is a little low number of ports
IMHO.

_________________
/Harry

[SOLD] µA1-C - 750GX 800MHz - 512MB - Antec Aria case

Avatar by HNL_DK!

@Samwel

If the user wants more stuff he should use off-the-shelf PC parts.

feature creep is the first enemy to battle in a design like this, and if you keep adding simple bits and pieces all over it ain't gonna be either on time or within price range.

So PLEASE reconsider whether what you need is USB, Sound and SATA inside the box or soldered to the board. If soldered-on CPU is such a MAJOR mental obstacle, why do we have the opposite feelings toward sound, SATA and USB?

Please discuss that while I go enjoy life for 5 hours

_________________
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.

@olegil

That looks nice, maybe you should do them an email.

Btw. i have done a search on google with "PowerQUICC Mini-ITX powerpc".
Oh, and i found that part from linux yesterday but i was not sure for
what you where looking for, i am just a simple user.

Edit: changed the search to + powerpc.

Edit2: hmm, not so good that search, i am trying "e300 PowerPC".

Last edited by Dirk-B on 04-Feb-2006 at 11:14 AM.
Last edited by Dirk-B on 04-Feb-2006 at 09:48 AM.

_________________
A1G3-SE + OS4.1 u1 iso (x2)

@Dirk-B

Man, the more i read abouth those SoC's the more i
think that we are missing the boat here if we do not
follow that pad.

Here is a link to the ibm socs:
http://www-128.ibm.com/developerworks/views/power/libraryview.jsp?search_by=SoC+drawer:

If you look at the classic Amiga with al those fine chips
on the board for every task lets say, well, all i can say
is that the next evolution is to put al those chips in 1 chip
and add some slots for expanding. Its like i saw it from the
first time i saw that Jeri Ellsworth C=64 games joystick.

But who is going to take the leaderchip here and organise
the needed roadmap for that. If we can go that way for our
low-end Amiga stuff then we have something in our hands.

Edit: changed the link.

Last edited by Dirk-B on 04-Feb-2006 at 12:21 PM.

_________________
A1G3-SE + OS4.1 u1 iso (x2)

@olegil

Quote:

Aha! Now we're honing in on the ####: http://www.silicontkx.com/GP2PRO.htm

Well, there is a pdf for this board:
http://www.silicontkx.com/pdfs/GP2PRO1.pdf

And i found a little interesting thing there:
"Open architecture : Uboot- no bios, no hidden APIs"

So i guess we can say that this is the best thing to
find for the moment, maybe it is the ****** device.

_________________
A1G3-SE + OS4.1 u1 iso (x2)

@Dirk-B

I'm fairly certain it's not the mystery device

I have some ideas regarding organisation structure, but for now I would be satisfied if a distributor would post here saying "you build it, I'll sell it". Because then we can start to worry about things like OS4 license etc

The way I figure it, this thing should pay for direct development cost in about 200 boards. which means OS porting, purchase of eval boards, design work, prototype run and the production/distribution cost of the first 200 units should be calculated into the price of those 200 units. In addition, OS4 would need to be bundled with it. Any units sold above that 200 unit line would make a profit, some of which should be dedicated to making a (or several) higher-end version(s), the rest split up among the team.

But I figure we need another 2-3 people involved: 1 distributor, 1-2 hardware guys (I'm game). Is there noone else mad enough to do stuff like this? Worst case, we don't get a working product. Very unlikely, the DDR interface is tricky but definetly doable. Second worst case we have spent a few days discussing ideas without getting a license, which I honestly can't see why would happen. (Then we need to discuss if MOS, AROS and/or Linux, BSD, Solaris are plausible options.) Best case we make a new Amiga!

Wouldn't that be great?

_________________
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.

@olegil

Hehe, well lets cal it a mysterious device then.

Like i could see in the other tread : OS4 hardware ,
there maybe be someone who is interested:

http://amigaworld.net/modules/newbb/viewtopic.php?mode=viewtopic&topic_id=17354&forum=33&start=100&viewmode=flat&order=0#261502

_________________
A1G3-SE + OS4.1 u1 iso (x2)

@olegil
well as long as you are involved with the hardware part, set up a paypal account and estimate the costs for a developer board. Iam sure there will be a few people helping out with a few dollars of sponsoring to get that going. as long as you keep the people informed and post news & updates at regular intervals that should be doable.

@IceDragon

Yes, that would be a good idea to do some sponsoring via a paypal account.

So next thing we need is a price for our first board. Or am i going to fast.

Last edited by Dirk-B on 05-Feb-2006 at 09:53 AM.

_________________
A1G3-SE + OS4.1 u1 iso (x2)

@IceDragon

I don't want to do this as a hobby project, I wanna do it as a professional product.

The GP2-PRO costs $495 and the eval board for the PCI - PCI bridge I mentioned (PLX6150 for &&MHz on one side and 33/66MHz on the other) costs $495. I would need 1 of each, and that would go to Hyperion for OS4 development. I can pay for that from my own pocket. What I need is help fund a 30000 USD production run (or at the very least 4 $7500 runs), there just ain't no way you sponsor that through PayPal. I need a dedicated distributor, then we can discuss finances. I have the contacts I need for purcasing PCBs and getting them soldered in units of less than about 10-20, but a board like this needs a professional production house, which means a professional operation.

Hardware wise, there are enough application notes from Freescale and PLX to actually make this design, now that I know how PCI interrupts work (found a good explanation in a secret PLX appnote ).

If someone promised to do the distribution and licensing (where are Gunne, Andreas, Rose, Justin and the guys at Datakompaniet when you need them, eh?) I would start working on the design yesterday. This would mean putting my own PCIe board on hold, but I would still like to make the PCI to PCIe bridge so we can use PCIe cards in our amigas (Mediator with PCIe Radeon, anyone?

So until a distributor pops up, this is as far as we get. The block chart is ready, I could whip up something in Gimp in 4 minutes to get ahead of the competition, as it were

The brainstorming was fun, guys

Now it's time to figure out the economics, which is never as much fun


Edit:
Prototype cost estimates (each, 5-10 unit batch):
SoC: $150
Bridge: $12
PCB: $120
GbE phy: $30
Connectors and smaller components: $50
Mounting: $200-300

Production cost estimates (each, 100 unit batch):
SoC: $120
Bridge: $12
PCB: $30
GbE phy: $20
Connectors and smaller components: $30
Mounting: $40

Rough estimates, I know. But it's not extremely far off. I've already gotten a price quote on the PCB from a Norwegian manufaturer, so the PCB costs are very real

I'll get some price quotes on the SoC and the DDR and PCI connectors tomorrow.

Last edited by olegil on 05-Feb-2006 at 10:26 AM.

_________________
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.

@Dirk-B
well the first thing would be to calculate the costs to get hw-development going i would guess. That would include the evaluation/dev-board, and everything else olegil needs to "work" with it. That would be an initial assessment of the starting (!) costs. Then we look who would be interested in that "venture", olegil said who is needed. After the "team" is assembled, the approach should be discussed and agreed upon in the team. Then you could get the paypal account setup and hoepfully soon start work on a first test-board. Etc...

@olegil
well then of course it's a bit different. Sorry if i misunderstood things.
Why don't you try talking toJens Schoenfeld from individual Computers ? He has quite some knowledge and contacts in that area (production, distributors, etc) and might be interested in such a project or maybe even glad to point out some help.

@IceDragon

I have the contacts (production wise), I don't have the funding. I do this for a living, remember?

Jens has his own product line and he has a bunch of sub-distributors. I need a distribution partner who can handle everything once the prototype run has been finished to everyones satisfaction. Between here and that point, I can handle it. But licensing and production runs is politics and banking, which I am not willing to deal with

Last edited by olegil on 05-Feb-2006 at 10:29 AM.

_________________
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.

@olegil

Sounds quite exciting !!

_________________
******************************************
I dont suffer from Insanity - I enjoy it

******************************************

@The_Editor

It isn't exciting. It's hard work designing, licensing and financing such an operation. So I understand it if no distributor wants in, but I'll still think it's a bit sad (btw, I only need financing on the production run itself, and only once the prototype is working).

_________________
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.

@olegil

Licensing first, next financing and then designing. You will not find any
distributor nor finance without proper contract(s) for AmigaOS4 (if this is the target).

I agree the first point will be hard work.

Bye,
TMTisFree

_________________
The engineering approach to our non-problems: "build a better washer".
The scientific approach to our non-problems: "find a new energy source".
The environmentalist approach to our non-problems: "stop washing your shirts".

@olegil

Quote:

So PLEASE reconsider whether what you need is USB, Sound and SATA inside the box or soldered to the board. If soldered-on CPU is such a MAJOR mental obstacle, why do we have the opposite feelings toward sound, SATA and USB?

We have drivers to SATA and sound. So why not with PCI card?

My personal opinion is, that if one controller circuit broke, it is more nice to change one card than whole mobo. At least in amiga-situation, where supply of new board is not granted.

Maybe USB can be onboard ? Alltough there is boards in market wich has USB-Firewire-SATA in one PCI card. Nice idea allso, alltough we dont have drivers fo FW.

Quote:

Or an even better idea: put in a PCI to PCI bridge, with PCI1 66MHz 1 slot (AGP connector) and PCI2 66MHz 1 slot + bridge and PCI3 from bridge to 1-2 33MHz slots. So you could have a 33MHz sound card and a 66MHz SATA controller like the SI3114 would still run at full speed. Example: PCI 6152-CC66BC from PLX Tech.

Sounds good . Personally I think, that PCI cards are good and acceptabel, if you have one or two slots free for future expansions :)

@Samwel

Quote:

$500 is a REALLY EXPENSIVE entry level machine! They need to have a complete Amy'05 board+OS4+case & stuff ready for less than $350 to atract the masses. And even then it will need several millions spent on PR. That may not even be enough as we have no important software available yet, like a office package and a browser.

If someone can spend millions in PR they can surely get a browser and an office pachage done.

_________________
My photos on Flickr

@samwel
Quote:

And even then it will need several millions spent on PR.

A simple webpage should suffice for this market. Selling Amiga's is preaching to the choir, not converting the masses. 1/8 of the potential market is probably reading this very thread. Most who would want one, already know they do.

_________________
"If you want to tell people the truth, make them laugh, otherwise they'll kill you." - Oscar Wilde

@TMTisFree

Eh, if you read what I wrote I said I would do the design, what I need is for someone to do licensing, financing and distribution. Seeing as the financing and distribution very likely needs to be in place in order to get a deal with AI for the license, it would be a bit silly to start on the licensing before the financing and distribution is in place.

So again, for this to happen I need to find someone willing to broker a deal with 1: Amiga Inc, and 2: a bank. No money would leave the bank until the prototype works, but at this point Amiga Inc needs to have said yes, which means the bank probably needs to have said yes as well.

_________________
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.