@fisk

Quote:

Ok, that makes sense. If the board already exists it's easier to use that. Still you would need to add at least USB2.0 on it. From the picture I can see it still uses the famous VIA southbridge (shudder).

I believe it's actually a VIA 8231, not the old 686B, and the 8231 seems to work okay. The Pegasos guys don't appear to have had any serious bother with it. I still wish they'd used the Ali 1535+, though, if only to keep the letters 'VIA' off the spec sheet.

But the 8231 gives the advantage of minimising the required changes to U-Boot and any drivers that hit the South Bridge, such as a1ide.device, and the Articia-P contains lots of southbridge-type functions (interrupt control, DMA, and the like) so it's very possible the 8231 is just there to act as an I/O device rather than an integral part of the chipset.

Quote:

Any A1-XC will be fine, I guess, at least if it's more advanced than the current models and maybe a bit cheaper I would not count on the latter one, though

No, nor me. I wouldn't expect to get all that much change from £800 for an XC with a G4 processor card.

_________________
Who do you serve, and who do you trust? - Galen

@DrBombcrater/All

WARMNING THIS POST IS JUST WHIMSY AND CANNOT BE TAKEN SERIOUSLY.

Ever wanted to produce your own computer generated movie with real life rendering?

Try a double G5 board in a XC and add in four 64 bit PCI-X based PPC cpu cards (already avaliable from teron designers) -- a possible 6 CPUs in a single box.

Real 3d already uses distributed rendering, the Articia P looks after througput and compositing. AOS4 if not now but soon will be able to cope with task sharing across muliiple CPUs and probably a lot more in the end.

What a box and even at premium prices this plugtogether would be so much cheaper than anything on the market, I mean what do you compare such a mother to?

Witho OS4, given some improvements under the hood, and its small foot print and UBOOTs ability to boot remote machinces, even I think I could set such a thing up and I'm a techno-idiot.

Mind you a whole bunch of Micro-A1-I on edge connectors would also be a powerful brew.

If we can get a few commericla customers to bring the HW prices down, we would become unstoppablke -- I kid you not. o-one could offer a single OS with so much HW punch, -- and to top it all AMIGA the damn nicest OS ever devisied.

_________________
Greg Schofield, Perth Australia

@DrBombcrater

Quote:

DrBombcrater wrote: Quote: As much as it hurts to watch this thing slowly hatch, why rush things just as the egg is cracking open, moving nests is not advisable, Well, there are undoubted advantages to that approach, but lots of potential pitfalls as well. The obvious one being we get what Eyetech (or more correctly, MAI) chooses to give us. Right now this amounts to one not very expandable motherboard, and that must be hurting the market to some extent.

Which market? The bleeding edge speed demons who want the beefiest, biggest, expandablest machine there is? Those people are chasing PCs. The number of people with the money to go that way on Amiga must be pretty small. I keep reading all the "when I get enough money to buy a uA1 ...." so we're already excluding people on price. Why go up? How much more would you have to charge to chase that tiny market..

Then there are the appliance makers. Since there aren't that many apps for desktop at present, appliance makers seems like a better bet. They are into putting together their own apps and what they want is a small and cheap board that runs quick and cool. Probably a smaller and less featured board would have a bigger market.

Quote:

Some people are going to keep their money in their pocket because the uA1 simply doesn't suit their needs.

Exactly. But where is the market. Do you have the studies that show where the dollars are? Or are you just asking for what YOU want?

@FluffyMcDeath

You have to remember where Eyetech stands in all of this -- their risks are very reduced to the ordering and pricing regime they use.

They can bring out small numbers of specialist boards at high prices with minmal financial exposure. Unlike an Apple run they can just order a few hundreed a tiime and yes these will always be expensive.

There are any number of businesses which could benefit from such flexible high end designs where money is no problem and PCs are not a favoured solution.

As for market studies and i have been involved with a few in the distant past - they can be next to useless (I have been involved with massaging figures with so-called reputioble companies) my advice to anyone, is ask a lot of people, think very carefully and use common sense. Big companies pay for expenbsive market research in order to shift blame when things go wrong, and prove their skills when things are OK -- its mostly just a game.

_________________
Greg Schofield, Perth Australia

@fisk

Quote:

And much worse I contatcted Freescale regarding the FPU implementation and is not opcode compatible to the normale PPC FPU

Perhaps not too surprisingly, there is already a supported, official way of dealing with that problem on the Amiga: math#?.library. Sure calling a function is not as efficent as gcc inlining FP opcodes, but then the OS is supposed to provide hardware abstraction and stability. If you want maximum speed, that's what #ifdef and inline asm are for

@Atheist

Quote:

Atheist wrote: Hi DrBombcrater, Comes down to what I've been saying all along, ibum threw CPU's onto the market and said "Have fun boys",

Intel made the CPUs. IBM made a dippy little underpowered simple machine with underpowered dippy little Intel CPUs and dinky almost non-existant support chips.

Korean factories who made IBM machines made illegal copies on the side and sold them themselves. IBM desperately tried to stop them but in the end decided to give up and focus on the big machines where most of their money was.

Because the IBM clones were so dinky and dippy they were easy to clone and soon they were everywhere. (Plus they had the business mystique of being "IBM")

It was nothing to do with support. It was to do with cheapness, profit margin and size of market.

x86 are Intel.
PowerPC are IBM.

@DrBombcrater

Quote:

PCIe would be nice, no doubt about that, but it's probably just a marketing advantage right now rather than a technical one. Same for DDR2. There's no significant performance advantage over AGP and DDR, and currently AGP cards and DDR memory are easier and cheaper to source than PCIe and DDR2 stuff. Given the complete lack of any ATX boards in the Amiga market right now a good machine this year is more useful than a very good one next year.

Maybe so, but still. Someone has to be looking at the next hardware design, and the time to do that is now. If I had a chance to buy a board with PCI-Express (not PCI-X) and DDR2, USB 2.0, and SATA with OS4 bundled, I would jump all over it. Those are the the things dreams are made of, in our market. A machine like that would be able to run absolutely anything you wanted to throw at it. I say "if you build it, they will come."

OS4 may run just fine on todays A1 hardware, and it may not push the full capacity of what the hardware can do, but what what we want is choice, and to not be left stuck in the past again. The A1 and OS4 is still all new, but the hardware has to begin to evolve and advance forward. The time to start is now, IMO.

Last edited by Herewegoagain on 26-Jan-2005 at 12:35 PM.

@GregS

Quote:

GregS wrote: @FluffyMcDeath You have to remember where Eyetech stands in all of this -- their risks are very reduced to the ordering and pricing regime they use. They can bring out small numbers of specialist boards at high prices with minmal financial exposure. Unlike an Apple run they can just order a few hundreed a tiime and yes these will always be expensive.

Let's see. Direct materials in small lots. $500ea. Manufacturing small run, shipping 20%,, probably more. $1million development cost / 500 boards ... That's $2600 before dealer markup - Sure, $3000 minimum. Cool. Then a there'll be all the complaints about how expensive it is... (and it'll still be underpowered).

You want to build toys like that? Sure, they may be fun, but that isn't a business. To make your bread and butter you need to be selling thousands at a time . Then you have the luxury to divert resources to the monster machines.

@DrBombcrater

Quote:

And the unexpectedly impressive specs of MAI's Articia-P board, the Teron PX2

That remark has made me pretty curious! I've just spent the past half hour googling and I can't find these specs though (and ftp.mai.com seems to be empty?). Could you post a link? pretty please?

@FluffyMcDeath
Lets presume you are right lest call it $3000 for a single board -- ie every very expensive by anyone standards (I would estimate a considerable smaller price less than $1500 US).

Does it matter? It may not be a buisiness in the run of the mill business needs. However there are special conditions, like needing to run passively cooled but fast throuput machines, needing to step up publishing corordination through a single console machine, or near enough real time 3d rendering at motion picture quality (many others could be added, none being big itself ie a few machines at best probably just one).

In any of these specialist areas money is far less of a concern than getting the right machine for the job, indeed the comptetion may well not be PCs at all, but powerful mini and fullframe machines with much higher costs and maintence bills.

Now admittedly a PC linux solution is a modest financial overhead, but a PC NT solution well that is another matter, a lot depends on the whole financial picture not just the board costs. For instance a good solid Linux build running a fully optioned up CX baord for small server needs (like schools) given the throughput possiblities, the greater stability and the competing maintence costs of NT -- well it could be a very choice item.

The of course there a number of folls like me that want a really powerful, mini-server that I can control through lovely little OS. Hardly worth a production run, (not a lot of big sales here I grant you. The Amiga community at this momentg would probably absorb a first small run even at exceptionally high prices.

Then there is digital TV services and TV companies in general and movie companies in particular, and the ability to do on-the- fly blue screen drop ins and camera, dolly, and robot digiotal stand-in control, along with editing notaion and the rest. Try doing any of this with a PC and six screaming fans.

In short, yes you are right, but vyou are mostly wrong. Eyetech would not make a great deal of money out of it at these high prices but it would not lose any either (that is really important). Remneber very little of this is done in-house, the financial risks are spead over a number of different comapnies all doing their own thing as well as this -- it is actually a really good company structure.

It is a srtucture ably designed to capture isolated niche markets, bring new moderls tgo these markets and slowly expand at the top end without high financial risk. I know of nothing else in this position that also offers the possiblity of compatible mass ordered commerical boards and cionsumer boards which can benefit from each such deal.

Eyetetch and OS4 get enormopurs potential market coverage out of this and very little financial exposure -- it is like every lesson learnt from the collapse of BEos has been taken on board (hint hint).

Low risk, constant design development, mulitple markets, full model range, huge reoom for market expansion. The reverse of Apple's strategy.

Last edited by GregS on 25-Jan-2005 at 04:26 AM.
Last edited by GregS on 25-Jan-2005 at 04:25 AM.
Last edited by GregS on 25-Jan-2005 at 04:24 AM.

_________________
Greg Schofield, Perth Australia

@CodeSmith

Quote:

That remark has made me pretty curious! I've just spent the past half hour googling and I can't find these specs though (and ftp.mai.com seems to be empty?). Could you post a link? pretty please?

Same here, I could not find anything on the Teron PX2 and would like a look at it.

I think everyone has some valid points here, and my personal opinion is that eyetech have been going about the hardware strategy the wrong way, though I think they made a step in the right direction with the current boards.

MAI having a board at that is ready to go is one thing, there are certainly many advantages with that, however with all the problems the A1 had with sound, DMA and so on, not to mention the difficulty with getting the right RAM, I find that trusting thier design again may be a little optomistic.Of course, maybe they have learnt and won't make the same mistake again, I guess in time with the current boards we will find out.

The problem at the moment, other than price, is that Eyetech have been all over the shop. Fist a desktop board, then the micros, and very little overlap of both these being available and thereby creating long delays for those consumers who want one solution or the other.

Eyetech, to be fair is trying to keep the users (us) happy, whilst also trying to develop a board for commercial use to actually make money on (not to mention make the boards cheap to keep us quiet and pay for more development) which is a hard task. I think the XC should be along the the same lines as the current micro boards.

If it is at all actually possible( given what is currently available in the chip market) I would make a board with the processor, USB, firewire, gigbyte ethernet, the cheapest sound you can get (along the lines of a SB16 in quality) and a cheap video solution...maybe a ATI 7500 or a 9200 LE, and lets not forget cheap but high speed RAM. I would go even cheaper than that but you do want the resolution capacity and as OS's are heading towards a 3d system you want at least something there that can handle a bit of 3d in the hardware.I would also try for a RAID controller on board as well.In other words, everything you need to hand high dataflow networking/serving. As has been mentioned, I would not bother with parrallel ports, ps2 standard serial ports or a joystick connector at all.I thiink it is high time we left this stuff behind.I am not sure what you need for diskless operation but have that on there too.

The last thing I would have on this board is the highest speed/bandwidth riser card like solution to allow for up to 5 PCIx/PCIe (take your personal preferance here) to allow for a full desktop solution.I would stuff all of this on the smallest board I possibly could, and if it was possible I would have it support G3-G5 processors, but as that is a pipedream have a board for each (or one for g3-g4 and one for G5)

Now I don't know if this is technicly possible, but I do not see why it would not be, or if it could get to the same kind of price bracket we have seen,but thats the direction I would go.I believe we DO need a g5 solution as well, the reasons for whitch I will go into later.The one extra thing I would have on the G5 board is support for 2 G5's on board.Of course, I would also have a several cases design with the branding prominantly on them.The brand exposure is an essential part that I think we are missing.

The advantages of this kind of design should be pretty self evident: the one board could cater for thos wanting a server solution, kiosk, workstation, office workstation, and have the capacity for expandability for those that want a decent graphics card, sound and other add ons. The costs involved should not change the price of the board too much, as Alan Redhouse stated the cost of the components is around 20% of the board, and so is the cost of fabrication.I dont see these costs being much higher.

Now thaqt we have these board that cover the full range of uses what next?We need to get costs of these boards down to take advantage of scale.As much as I hate to admit it, BBRV did have some good ideas. One of these is to push for and really get linux support going for his boards.
People may be reluctant to go to such expense with OS4 and the limited software available on it, but if you have a very well supported linux distribution you can push that aspect to the many corporations and even governments looking to change away from windows, coupled with the low power and other advantages of the PPC cpu have a real product to sell in large numbers.The last advantage is that we would have a backup in the case OS4 does fail for whatever reason.(Who knows what KMOS/Amiga Inc is doing, and you never know, a meteor could always fall on our good Hyperion budies...)Of course, securing the financial side of our hardware would give Hyperion a bit more time to get OS 4.0 right and take some pressure off them, which is a good thing :)

Alot of people will object and/or dismiss the G5 idea. But we have to face it, the world is moving towars 64-bit and we would be foolish not to go that way as well. Not to mention that having a high end solution now would certainly help us move to 64 bit, attract customers interested in that market, and allow us to have a mature hardware platform for when we do go to 64 bit, so we can avoid a repeat of the A1-XEproblems when the time comes.

The last thing I would do involves DE and having that native to OS4.But that depends entirely on how good intent2 actually is and if it lives up to what it has been advertised to be.

Last edited by IonMane on 25-Jan-2005 at 04:59 AM.
Last edited by IonMane on 25-Jan-2005 at 04:56 AM.

_________________

If A1XC will be microATX, it will only allow a total of 4 slots.. according to microATX specs...

so to all those dreaming about 5 PCI slots + AGP, forget IT...

simply look at pegasos 1 and 2 and other PC microATX mobo...

@DrBombcrater

Quote:

I believe it's actually a VIA 8231, not the old 686B, and the 8231 seems to work okay. The Pegasos guys don't appear to have had any serious bother with it. I still wish they'd used the Ali 1535+, though, if only to keep the letters 'VIA' off the spec sheet. But the 8231 gives the advantage of minimising the required changes to U-Boot and any drivers that hit the South Bridge, such as a1ide.device, and the Articia-P contains lots of southbridge-type functions (interrupt control, DMA, and the like) so it's very possible the 8231 is just there to act as an I/O device rather than an integral part of the chipset.

Another good point of you. If the Articia already has most of the SB stuff, then we could also completly ditch the SB and just connect some PCI periphal chips directly to the second PCI bus (AP can be configured for 2 PCI and 1 AGP). Only if the 8231 is cheaper it makes sense to use that. But I thing today you would only need SATA/PATA chip and USB2.0 and maybe a serial port for debugging...

Quote:

No, nor me. I wouldn't expect to get all that much change from £800 for an XC with a G4 processor card.

Yeah that is quite sade. but I thought a bit about it during my walk to work and it seems the price must also include additional cost for possible need of exchanging boards if there appears a HW defect after shipping to customers.

Any cheaper motherboard for us OS4 users would be very welcome though. And I believe it's only possible using the new SoC solutions of Freescale (or AMCC (former IBM 4xx parts)). As Eric_S said the 8641 would also be in a good time frame, but the bPlan/Genesi guys also want to use that chip for a board and they seem to have good contacts with Freescale; so they have an advantage. Best solution would be to cooperate and allow OS4 to run on their HW. But sadly we have this cold war with them

And of course we need good software, too. So maybe I will just get back to my software plans and hope someone else will do the right thing for the hardware

@fisk

Quote:

Don't worry, I did not want to flame on the Athlon (better than P4), I just wnated to show that we do not need much GHz for OS4

The market has changed i.e. the acceptance of Pentium M and PowerNow/SpeedStep enabled desktop A64s/PIVs. With A64-CG, 3 MPEG2 @SVCD level doesn?t trigger clock speeds beyond 800Mhz.

Last edited by Hammer on 25-Jan-2005 at 07:22 AM.

_________________
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB
Amiga 1200 (Rev 1D1, KS 3.2, PiStorm32lite/RPi 4B 4GB/Emu68)
Amiga 500 (Rev 6A, KS 3.2, PiStorm/RPi 3a/Emu68)

@IonMane

Quote:

Same here, I could not find anything on the Teron PX2 and would like a look at it.

You can look at a picture of the PX2 v1 here. I have quite detailed specs for this board, which come from a very plausible source and tally 100% with the picture. I'm not prepared to say any more about that source.

The picture is from an MAI/AtumTech powerpoint presentation that popped up on Atum's now defunct web site some time ago. There's a news post on AmigaRulez that referenced the presentation and mentions the PX2, but I can't read whatever language its in and didn't make the connection until very recently when I got more information on the PX2.

The board in the picture is the Teron PX2 v1, I'm totally certain of that. The specifications for it are:

* Articia P North Bridge with 200MHz FSB
* G3/G4 CPU via 300-pin Megarray CPU socket
* VIA VT8231 South Bridge
* 1 4x AGP socket
* 2 64-bit PCI slots
* 2 64-bit PCI-X slots, clock speed unknown but probably 100MHz.
* 3 32-bit PCI slots.
* 3Com 3c920 100Mbit ethernet
* 4 184-pin DDR memory slots, max DDR400
* AC97 audio via the VT8231

The v1 is just a prototype and has been pictured in public so I don't mind discussing it, but the later v2 revison seems to be the real deal and will probably ship as the A1-XC. It's still under wraps and I'm therefor rather reluctant to talk about it in any detail. Suffice to say production cost reasons mean it is cut-down from the v1, but not much, and anyone who's interested in a powerful, expandable A1 will love it.

Yes, I know I'm waxing mysterious here, but there are good reaons for that.

_________________
Who do you serve, and who do you trust? - Galen

@DrBombcrater

Grr!

_________________

@DrBombcrater

Quote:

DrBombcrater wrote: It's still under wraps and I'm therefor rather reluctant to talk about it in any detail. Suffice to say production cost reasons mean it is cut-down from the v1, but not much, and anyone who's interested in a powerful, expandable A1 will love it.

Sounds great

@DrBombcrater

for the news

And for the informations.

Jerry

_________________
- AOS has been ported to ex-86 ! It's called AROS and WinUAE... Or E-UAE on Linux !

- A1XE-G4 up and runing with:
512MB Ram / 200GB and 80GB HardDisks on Sii0680.
AOS4 Final Update / AmiZilla 0.1 Alpha

@DrBombcrater

Quote:

I have quite detailed specs for this board, which come from a very plausible source and tally 100% with the picture. I'm not prepared to say any more about that source.

Allright! now THAT is a cool amiga rumour. Those specs are pretty neat

Thanks for the info, and please thank your source for us. As long as the info is reasonably accurate, I don't mind not knowing who it comes from.

@DrBombcrater

Wow! If that's true, it will be about 5 times better than the current A1XE specifications, and the A1XC would then be a NEW full ATX-motherboard, and
not a microATX board then, based on the Teron PX Revision 2 board as seen
on the screenshot. NOW we're talking hi-tech!

Eyetech! Time to contact those MAI Logic guys to make these A1XCs based on the Teron PX Revision 2 board, and i'm sure that you will find much better chips than those from VIA! Fuse VIA! Find chips from better companies!

This is just great rumours! I love it!