[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [openppc] Pegasos PPC Motherboard



> (and anyway,
> i have no money right now for buying one :)))

And there's no way I can justify another computer !!


> I have seen on one of the rumor sites, that one of the apples G5 developper
> board (the second batch) is based on hypertransport, but now, is it true ?
>
> Well, my understanding is that the G5 includes the memory subsystem and the
> rapidio/HT bus. Don't know how AGP will work with this, but i guess it will be
> an independent HT/rapidio device, with an AGP bridge or something such.
>
> If this is true, then the bus choosen will be important once the graphic chip
> maker will include the HT or 3GIO port directly, instead of using AGP.
>
> Also, we have not to forget 3GIO, where AMD is a participant also, ...

AFAIK the idea with AGP is to have a RapidIO/HT <-> AGP bridge, otherwise you've got to wait for
dedicated RapidIO / HT graphics chip sets (they same applies to PCI). As to Apple's next move, who knows
- any one any idea when they might release G5 stuff?? (or when Motorola might for that matter??). I
wouldn't put it passed Apple to use HT and use a chip set from (say) ATI, which hasn't ben anounced
yet....

As to 3GIO, AMD says it's complementary to HT, so we may yet see it. AMD were certainly talking about HT
<-> 3GIO bridges, so both could be used inside the same box - that leads on to other interesting places
(once 3GIO stuff is available).

I'd also refer folk to http://www.tundra.com/page.cfm?TREE_ID=101039 (Cheers Sven, who originally put the
link up)
It looks like a very promising chip, though it's not out yet. Could we see this in the next generation of
PPC based designs. Could this be the NorthBridge we've been waiting for??


> > P.S. Anybody able to attend Amiga-2001 in Cologne?? and report back??
>
> Not me, it is a bit far, and i wouldn't be able to justify the deplacement
> anyway.

Like wise :-(

Alasdair

--
To unsubscribe from openppc mailing list please visit http://www.opencores.org/mailinglists.shtml