Because it doesn't exist.



You realise that'd be a $1500 motherboard, right? The Pegasos G5 was canceled because it was unaffordable. There is a good reason that those Apple systems cost $3000.

Fixstars are probably selling the PowerStation box at a significant loss (we know because we designed the motherboard). This is just the reality of the 970 pricing and the incredibly complex board design required to support it (many weird chips, braindead design decisions by the Northbridge and CPU designers..).

is it just me, or have you, guys, been almost single-handedly driving the ppc consumer market until recently? : )


i'm curious to know what you found wrong with the cpu design.

It was a contract with IBM. When it turned out the motherboard would cost so much and there was no opportunity for price breaks and they reneged on a few things, we stopped.

IBM own the board design and took it to the next step, but didn't fix any of the expenses.



The CPU is fine.. just too expensive. The Northbridge is oddly packaged and needs to be placed on the underside of the motherboard to reduce the complicated pin routing. It needs a HT bridge and a HT southbridge (expensive option, and there are only a couple on the market). All in all, it is just a very costly design obviously not meant for anything but servers where service fees make up profits and not hardware.

The P1022 doesn't exist either ...

Markos,

maybe you should also post on:
ArsTechnica Forum
RealWorldTech Forum
IBM CELL/BE Forums
Beyond3D Forum
Apple Mailing List

News on Wired would be cool, too.

At least there is a firm roadmap for sampling fairly soon, but 6 months of waiting could be filled with 6 months of development on the MPC8610 board.

Why still not use the P4080? Well, simply put right now, not only does it not exist, but what are you going to do with 8 cores, 8 gigabit ethernet and no SATA or video or audio? It's a worse choice than the MPC8640D in this regard.

WHAAAAT? And we knew nothing about that until NOW?
Hard to believe, until I red some techincal details from you.

First shocked, now impressed (yet again).

Bill posted the information about a year or two ago?

PowerStation is the final result of the TetraPOWER motherboard, with whatever changes they needed to make to get it to work. We got as far as PCB production and initial prototype when it got to the point that IBM said they would not buy any to sell to developers (I guess they wrangled Terrasoft into doing this) and that they would never give us the prices needed to make it an affordable developer box.

The BOM cost for a 1.5GHz quad-core board was over $900 (reaching 750 EUR at the time!) including processors for the complex dual-sided PCB. For >2GHz it hit $1500. Genesi would not sell a single unit at that price. Given that it had to use expensive RAM, expensive disks, weird cases (BTX pretty much got abandoned halfway through).. we would have to sell systems at the same cost as an Apple G5 (nearing $3000).. which defeated the whole object of the project.

People still ask about it and we point them at Fixstars.. or the standard route of buying a G5 on eBay. Part of the beauty of the Pegasos and Open Desktop Workstation was a $799 price point for a complete system - even that was a little over expensive for most people (consider a dual-core Mac Mini is around $599)

Matt,

had Genesi ever talked to PA Semi at any time? If yes, how did it go?

Yes, under NDA :)

mmh... PCIe (and a riser card) 'd be nice if a valid openfirmware-like (or u-boot devtree) is available to really use it with different add-on cards.

A PCI Express X8 to X16 Extender could be useful to use a wide range of graphic adaptors.

It certainly went nowhere because PA Semi and its wonderful PPC design are dead since the acquisition by Apple.

Already send the mail with my interest a few days ago. The 8610 is the chip of my choice. I would like a board that would have become the Efika 8610. Small, low energy and still affordable. But indeed, it will be not as cheap as ARM (or Atom for that matter, but who wants Atom anyway;-)). I am quite optimistic an 8610 board for a reasonable price will find a couple of hundred customers. If it is designed by bplan it probably will work well. But the word has to be spread to sell the boards.

More updates here:

http://www.codex.gr/index.php?pageID=&blogItem=64

allow me to comment on the next:


i'm not sure you could say that just based on the feedback emails - this is your core audience (or part thereof), but that does not mean that's your entire market.

a couple of considerations:

* my impression is that compilers still love the ppc architecture - with all the popularity ARM currently enjoys, it's yet to reach the level of compiler support that ppc (well, book e, at least) has had for long now. it's because the ISA was designed as the sane middle-ground architecture - nothing 'common' is prohibitively complex to implement on it (you need cheap dp fp - sure. you want sane SIMD - got it).

* how many non-x86 ISAs can boast with such excellent x86-borrowed pheripheral support? yeah, it's been mainly a matter or proprietary emu tech, but the plafrom at least has that. how many of us are running PC pheripheral cards in our obscure ppc boards - boards that 90% of the computing world has not even heard of? of all 'non-comformant' architectures, ppc has the biggest intersection with commodity tech. plus, we can build the bleeding edge linux kernels at the effort of setting a couple of build switches (and the occasional patch).

i know everybody 'prudent' is going ARM these days, and soon it will be a ubiquitous platform (if something that's already in gazillion of CE devices can become any more ubiquitous, that is), and that IBM cannot be arsed to move their little finger to put ppc back on the consumer-maket map (outside of a couple of hundred of millions of game consoles), but i thought the whole point of your effort was to cater to the ppc geek crowd, not to seek the greenest pastures (you could just as well re-sell atom boards to that end : )