Boston Linux & Unix (BLU) Home | Calendar | Mail Lists | List Archives | Desktop SIG | Hardware Hacking SIG
Wiki | Flickr | PicasaWeb | Video | Maps & Directions | Installfests | Keysignings
Linux Cafe | Meeting Notes | Blog | Linux Links | Bling | About BLU

BLU Discuss list archive


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

[Discuss] AMD FX-8120 update



On Mon, Mar 5, 2012 at 11:53 PM, Bill Ricker <bill.n1vux at gmail.com> wrote:
>> I'd be happy to explain the high-level details of the architecture to
>> anyone that cares.
>
>
> So is the Northbridge effectively half on-die and half off-chip ?

Yes: slide #28 illustrates this to some extent. The "Valencia" die has
the part of the northbridge with the crossbar, L3 cache and the DRAM
controller. The SR56x0 "Northbridge expansion I/O" chip contains the
I/O controller and the PCIe I/O lanes (among a few other things). The
Valencia die and the SR56x0 are connected via a HyperTransport link.

Slide #28 is for a 2-socket server configuration -- for a single
processor desktop or server, there will obviously be only processor
one die at the top and the part numbers for the expansion Northbridge
and Southbridge may vary.

When we get to the upcoming AMD Fusion APUs for desktops and laptops
based on the Bulldozer core architecture (punch in 'AMD Trinity' into
your favourite search engine), the story is obviously different. The
entire Northbridge, i.e. both the memory controller part *and* the I/O
part, will be on the same die, along with the Bulldozer modules and an
AMD Radeon HD graphics controller. There are a few other differences
that I probably can't talk about yet :-)

-Shankar



BLU is a member of BostonUserGroups
BLU is a member of BostonUserGroups
We also thank MIT for the use of their facilities.

Valid HTML 4.01! Valid CSS!



Boston Linux & Unix / webmaster@blu.org