@Raffaele

But first of all...was Hyperion and MOS team interested to the thing?
Will they supporto it?

_________________
Simone"Tuxedo"Monsignori, Perugia, ITALY.

@tonyw

The X5000/40 is a full system costing over $3,000. And the Cyrus Plus motherboard-only option will still be upwards of about $2,000.

I think @Raffaele believes that if the Cyrus Plus motherboard were not designed and built by Varisys i.e. not using expensive person-hours from the west but rather leverage global economies and have the board designed and built somewhere in south-east Asia. There are many talented electronics engineers in the Phillipines, for example, that will work for $300 per day instead of lets say $1,500 - $2,000.

And there are many manufacturers throughout Shenzhen that would happily produce ~2000 motherboards for something like $100 a piece.

Of course there is more to bringing a product to market, such as prototyping, testing, hardware revisions, more testing, drivers, linux port validation, electronics certification (CE, FCC, etc).

I certainly believe it's possible to bring a Power ISA board to the market that is less of a compromise than an A1222, and is also less of an expense than a Cyrus Plus. It's all in the execution. What Trevor and co. have done is one way to do it, but it isn't the only way.
Unfortunately there wasn't that much "oxygen in the room" to begin with and I doubt there is a 1000+ user base for this "dream" board.

_________________
All the way, with 68k

@Raffaele

My specs:
https://communitycomputer.wordpress.com/about/techspec/

_________________
- KimmoK
// For freedom, for honor, for AMIGA
//
// Thing that I should find more time for: CC64 - 64bit Community Computer?

@Raffaele

And then what SS said "what OS do you plan to run on this new board?".

And let's not forget the time scale. It'll take at least a year to produce the hardware, but more like two years if you're keeping the costs down and you're collecting bounties and such. And then what, 6-12 months to port Morphs/AmigaOS to it?

So your "dream" board spec should be one that an Amiga user will want 3 years from now.

A little demoralising I know, but you'll need a strong constitution to survive in this game.

_________________
All the way, with 68k

@thread

I always liked the idea of the Dome microserver.
Adapting the system for desktop could be trivial for experts.

The creator sayd that the bill of material was around 600 dollars.

Btw, Tabor is alread a good starting point.

_________________
retired

Why not to try to involve the only company who is producing ppc chips, desktops and laptops and servers. They may be interested in supporting ppc market making a ppc machines not just servers.
IBM has all that and the power to make that.

@Raffaele

It would be cheaper and a lot more cost effective to do a port to an existing mobo.

_________________
Dammy

@Raffaele

The most logical CPU core right now is to me (as KimmoK has already hinted at) an e5500, but the most serious CPU core is obviously the e6500 for Altivec. A T1022 design with an AMD southbridge and a slot for graphics is the most obvious choice then, as this allows you to use also T2081.

I've harped on and on about this so many times now it's getting just depressing that noone else seems to agree with me.

My suggestion:
Don't go down this route without some SERIOUS technical knowledge.

_________________
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.

@Raffaele

I think this is a great idea, and please document each and every step, compromise and design decision made for everyone to see. That way it might serve as a guide/warning for all those in the future who believe it's an easy task and that they can do better.

Quote:

Raffaele wrote: Once we will have the first two or three solid iron prototypes then one of the prototypes will be sent to Hyperion and another prototype will be made available to MorphOS team in order to let them realize a first OS porting. Phase 4: We will choose the manufacturer who will reveal itself as more serious, and ask the price for a first batch of motherboards. Phase 5: Starting of a second bounty for manufacturing our own community motherboard

Why would anyone from either MorphOS or Hyperion dedicate the amount of time required to get an OS working on a prototype motherboard when there's no guarantee the board will ever get made? You need a commitment to build and release it to make it worthwhile, otherwise there's a risk they will spend months getting an OS to work on a board of which only 3 prototype units exist worldwide.

Also, how will you verify the prototypes? And who will do that work? Up until now it has involved porting Linux to the board so all the on-board peripherals can be tested with existing drivers or based on existing code. Do you know someone willing to do that work? Hiring someone to do it will be expensive... Very expensive.

Sincerely, the very best of luck with it. I know how much work goes into developing much simpler boards for embedded controllers, so I know it will be extremely difficult. If you somehow manage to hit all your goals, then brilliant - cheap, powerful hardware for OS4 and MorphOS is a good thing. And if it doesn't work out, at least there will be a story there for the inevitable times that people criticise the work gone into boards like the X5000.

_________________
RobTheNerd.com | InstallerGen | SMBMounter | Atoms-X

New mother board ppc with good FPGA for retro compatibility (aga,paula,etc...)

@Raffaele

Hey, Raffaele, I like this! Not only have you been presenting a grand idea, whether plausible or not is not up to me to tell: I'm qualified only for a bookkeeper, you have also given start to a healthy discussion. As always, there are some people who will jump on it and burn it down, but thankfully there are also some that will chime in with solid knowledge about the matter. You not being hindred by deep knowledge is taking the discussion into a direction hardly given a chance before.

As Albert Einstein seems to have said:
"When you never get lost you never discover a new way"
and:
"Great spirits have always encountered violent opposition from mediocre minds"
and:
"Imagination is more important then knowledge"

(AA seems to have provided the world with lots of really quoteable remarks)

So, go on!

OldFart

_________________
More then three levels of indigestion and you're scroomed!

@Raffaele

why not either support existing hardware or support ISA change?

and only a small thing... someone has to sign the bills in development phase, someone has to sign the bill with the chinese suppliers, if something goes wrong someone has to pay for it, to sell you must offer guarantee, you need logistics and so on. It is a business venture, not a home-brew project where some prototypes are developed for private purposes

The developers behind Apollo Project can support (propably) a affordable solution on Vampire because all development is done for free by enthusiasts. This would not be the case at your project.

Last edited by OlafS25 on 04-Nov-2015 at 11:04 AM.

Well, I'm not eager to jump to the doom and gloom phase in the initial stage. I like the idea and IF you can make it happen, I will support it if these criteria are met:

1. It has to be supported by all NG camps (though I think that because of the nature of things, AROS can take it's time).
2. It has to be considerably cheaper.
3. It won't come installed with any OS by default. This is to make sure it doesn't become associated with any camp.

Good luck! I will keep my eyes open

Last edited by Yasu on 04-Nov-2015 at 11:08 AM.

_________________
Amiga Forum - Sweden's best Amiga Magazine

My MorphOS Blog

"Free speech includes other peoples right to offend you."

@OlafS25

Quote:

OlafS25 wrote: @Raffaele why not either support existing hardware or support ISA change?

Because it is either discontinued (Mac), underpowered (AOnes clocked at 800 MHz) either expensive (X1000, X5000)...

About ISA change I don't know hidden projects of Hyperion and MorphOS Team. They are the only who can reveal about ongoing ISA changes. For any different approach we have AROS then.


Quote:

And only a small thing... someone has to sign the bills in development phase, someone has to sign the bill with the chinese suppliers, if something goes wrong someone has to pay for it, to sell you must offer guarantee, you need logistics and so on. It is a business venture, not a home-brew project where some prototypes are developed for private purposes The developers behind Apollo Project can support (propably) a affordable solution on Vampire because all development is done for free by enthusiasts. This would not be the case at your project.

We are far from the point to sign contracts and pay bills for manufacturing.

We are still in unknown land and explorating. First learn to walk, then to run.

Last edited by Raffaele on 04-Nov-2015 at 11:21 AM.

_________________
"When the Amiga came out, everyone [at Apple] was scared as hell." (J.L. Gassée, former CEO of Apple France and chief of devs of Mac II-fx, interviewed by Amazing Computing, Nov 1996).

@Raffaele

Quote:

I beg your pardon?

if im not mistaken, you have not even checked the site you linked to, what it is about? where is electronic circuitry prototyping mentioned in there? based on that, i have some doubts if you are seriously ready to invest any time, responsibility and effort in that kind of project. you would be managing peoples funds, remember, what would be your explanation after something went wrong? "beg your pardon"?

Last edited by wawa on 04-Nov-2015 at 11:25 AM.
Last edited by wawa on 04-Nov-2015 at 11:24 AM.

@Raffaele



Quote:

CPU APM Black Mamba dual core 64bit upto 1.2 GHz with NO Altivec

Quote:

CPU with no Altivec and with a anomalous FPU is not serious for me.

I'm confused now... Is Altivec required or not?


Quote:

Phase 2: Collect bounty for prototyping and make a business plan on how about this motherboard shoupd by produced by a China Manufacturer

What was Phase 1? I thing that working on a business plan should come before starting a bounty... You need funders to have an idea of what they are buying into.

Quote:

Phase 3: If we will reach enough money for prototyping then we will start searching a manufacturer.

If you don't have a manufacturer in mind then you don't know how much will be enough money. Find a designer/manufacturer needs to be part of Phase 1, along with the business plan. Bounty is still at least Phase 2. When did you search for and find a designer?

The designer really needs to know manufacturing capabilities before starting, so it doesn't make sense to search for a serious manufacturer after the designer is working. If the designer makes some assumptions, and later you find a manufacturer that isn't kidding around, then there could potentially be some redoing of design work based on that manufacturer's capabilities being different than the initial assumptions that were made. Let's know the capabilities and design rules first.

You can't just start a bounty, and then after defining that start looking for designers and manufacturers. You don't know how much you need that way.



I've been pouring over your kind of suggestion for over a decade. I have two finely tuned concept and spec, and most of a master degree in electrical/engineering to show for it. If I could choose certain things for the because of it, then it would be a single concept in both cases. But that would require a lot more CAD for the other case. And after all this I have no PCB design or even any schematic worth mentioning.

The 8641D (really the same thing as 8640D from your second post here) was part of that several years ago. I wouldn't consider it serious enough today. I forget if I considered it before or after the PA16T as found in X1000 was eaten by a fruity company. Both were long ago choices. I'm not familiar with Black Mamba CPU, I'll read up on it a bit, but I'm really interested in e6500 or better core type WITH altivec.

I'm very hopeful about the Linux/BSD PowerPC notebook project. I'm not sure how much faith to put into that yet, but they do at least have more of Phase 1 going on than you do.

_________________
All glory to the Hypnotoad!

@Raffaele
Hi like Olegil says:
"A T1022 design with an AMD southbridge and a slot for graphics is the most obvious choice then, as this allows you to use also T2081." Is in my opion the best way for a new design.

If you want it cheaper you can remove the southbridge and add 2 x1 PCI-E slots and 2xSATA connectors direct to the chip.

A prototype board will cost you approx. 3.000 €
With 3 revisions you need 10.000€ without the costs of the board-layout.

The tricky part of designing boards is the power supply for the chip (dc/dc converters) and the high speed signals DDR-Ram and PCI-E/ SATA.

Another way of doing a new board is to take a SOM (System on Module) and design your Carrier Board.
For example:
http://www.sintecs.eu/modules/SiMC-T10x
The advantage is that the design of the carrierboard is much easier and cheaper and you can plug in a pin compatible SOM.
Another benefit is that you can start your software development immediately on a stable system.

Produce these motherboards in China.....

One of the nice things about Acube's efforts are that they are made in Italy and they try to use EU suppliers for components..... Using components from EU based companies benefits the local supply chains.... And sometimes you cant really put a price on that....

Environmentally speaking it is also better for boards to be produced locally rather than shipped halfaway around the world in a container ship....

_________________
Amigaone X5000/20 4GB Radeon RX 550 Polaris 12 AmigaOS4.1 Final Edition Update 1
Amiga 1200 Workbench 3.1.4
Amiga CD32

@THREAD

A T1042 1.4 Ghz board for the low end.
And a T2080 for the high end.

Don't worry about pin compatibility as the first could be Mini itx and the second MATX.

And I have already approach two organizations I'd like to get involved.

The T10XX would, of course, be the first target.
Even though the costs will be similar, if its priced low enough it may be hard to undercut Tabor.

BUT, it would function MUCH better.


Edit - Freescale has posted the design files for the T1040RDB reference design board on their website.
I don't know if they are available to the general public or if you have to be registered with Freescale.
If the latter is the case, I'm willing to send a copy to anyone interested.

As a reference design, its had its issues, but it is on its second revision.
And it would provide a starting point.

Last edited by iggy on 04-Nov-2015 at 01:09 PM.
Last edited by iggy on 04-Nov-2015 at 01:04 PM.

@cdimauro

Quote:

First, you didn't ever took a look at the processors' data sheet to see if they are compatible with the motherboards specs that you proposed.

Well said!

@Raffaele

I know you mean well, but your proposal is not based on reality. Recent Freescale SoCs (10XX) would be much better suited for such low-cost board.
Note: Black Mamba is 32 bit. Both your SoCs of choice don´t support features you want.