Looks like they found that they had to respin the DDR2 memory
controller circuitry, because it wasn't performing too well. That's why
it's now slated to be released by June 6, 2006 (6-6-6, btw), instead of
earlier.

Geek.com Geek News - DDR2 issue causes Socket AM2 delay
http://www.geek.com/news/geeknews/20...0220034869.htm

Yousuf Khan

YKhan <(E-Mail Removed)> wrote:
: Looks like they found that they had to respin the DDR2 memory
: controller circuitry, because it wasn't performing too well. That's
: why it's now slated to be released by June 6, 2006 (6-6-6, btw),
: instead of earlier.
:
: Geek.com Geek News - DDR2 issue causes Socket AM2 delay
: http://www.geek.com/news/geeknews/20...0220034869.htm
:

Is this the replacement (ie, AM2) for socket 939? I ask because I would
like to build a new system within the next six months or so, and am a
little confused with all the changes of late. My Asus A7V333 is getting a
little long in the tooth, and since I do lots of Photoshop work and lots
of MP3 and video encoding, the thought of a dual-core AMD platform is
becoming more and more of "a need" for me. :-)

G7

Gary Seven wrote:
> Is this the replacement (ie, AM2) for socket 939? I ask because I would
> like to build a new system within the next six months or so, and am a
> little confused with all the changes of late. My Asus A7V333 is getting a
> little long in the tooth, and since I do lots of Photoshop work and lots
> of MP3 and video encoding, the thought of a dual-core AMD platform is
> becoming more and more of "a need" for me. :-)

Yes, it's socket that AMD is going to be using to support DDR2 (and
apparently also DDR3, but we'll have to wait to see that).

Yousuf Khan

On 20 Feb 2006 11:56:06 -0800, "YKhan" <(E-Mail Removed)> wrote:

>Looks like they found that they had to respin the DDR2 memory
>controller circuitry, because it wasn't performing too well. That's why
>it's now slated to be released by June 6, 2006 (6-6-6, btw), instead of
>earlier.
>
>Geek.com Geek News - DDR2 issue causes Socket AM2 delay
>http://www.geek.com/news/geeknews/20...0220034869.htm

Hmm, I wonder if this why they signed up for the Rambus license?... to get
some engineering help?

--
Rgds, George Macdonald

George Macdonald wrote:
> Hmm, I wonder if this why they signed up for the Rambus license?... to get
> some engineering help?
>

It's likely whatever problems they were having was based on their own
designs, so they'll have to fix their own bugs. Rambus might be there
for future revisions the DDR controller. Hell, it might even be there to
help out with the ZRAM controller. :-)

Yousuf Khan

Yousuf Khan <(E-Mail Removed)> wrote:
: George Macdonald wrote:
:: Hmm, I wonder if this why they signed up for the Rambus
:: license?... to get some engineering help?
::
:
: It's likely whatever problems they were having was based on their
: own designs, so they'll have to fix their own bugs. Rambus might be
: there for future revisions the DDR controller. Hell, it might even
: be there to help out with the ZRAM controller. :-)

Rambus? Is this the **same** Rambus from just a few years ago that was
running around sueing the pants off everybody and their grandmother? Why
o why in god's name would AMD want to get involved with those vultures?
:-(

G7

On Tue, 21 Feb 2006 22:46:31 +0100, "Gary Seven" <(E-Mail Removed)>
wrote:

>Yousuf Khan <(E-Mail Removed)> wrote:
>: George Macdonald wrote:
>:: Hmm, I wonder if this why they signed up for the Rambus
>:: license?... to get some engineering help?
>::
>:
>: It's likely whatever problems they were having was based on their
>: own designs, so they'll have to fix their own bugs. Rambus might be
>: there for future revisions the DDR controller. Hell, it might even
>: be there to help out with the ZRAM controller. :-)
>
>Rambus? Is this the **same** Rambus from just a few years ago that was
>running around sueing the pants off everybody and their grandmother? Why
>o why in god's name would AMD want to get involved with those vultures?
>:-(
>
>G7

AMD can't afford anyone/anything else.

AMD inked a 5 year patent license agreement with Rambus Inc. that grants
AMD a license to Rambus patents used in the design of DDR2, DDR3,
FB-DIMM, PCI Express and XDR controllers along with other current and
future high-speed memory and logic controller interfaces for $75
million.

Yousuf Khan wrote:
> George Macdonald wrote:
> > Hmm, I wonder if this why they signed up for the Rambus license?... to get
> > some engineering help?
> >
>
> It's likely whatever problems they were having was based on their own
> designs, so they'll have to fix their own bugs. Rambus might be there
> for future revisions the DDR controller.

I suspect it was for engineering help as George put it. A lot of guys
have used Rambus to design fast memory controllers/subsystems, since
they have lots of expertise in the area.

> Hell, it might even be there to
> help out with the ZRAM controller. :-)

Yousuf, you're a very funny person.

DK

On Tue, 21 Feb 2006 15:55:25 -0500, George Macdonald
<fammacd=!SPAM^(E-Mail Removed)> wrote:

>On 20 Feb 2006 11:56:06 -0800, "YKhan" <(E-Mail Removed)> wrote:
>
>>Looks like they found that they had to respin the DDR2 memory
>>controller circuitry, because it wasn't performing too well. That's why
>>it's now slated to be released by June 6, 2006 (6-6-6, btw), instead of
>>earlier.
>>
>>Geek.com Geek News - DDR2 issue causes Socket AM2 delay
>>http://www.geek.com/news/geeknews/20...0220034869.htm
>
>Hmm, I wonder if this why they signed up for the Rambus license?... to get
>some engineering help?

LOL! Interesting theory, but I really doubt it. My reading of it is
that it's just a performance-related issue with the layout of the
circuits. This sort of thing does happen, and while it's not
particularly good news for AMD, it sounds like something that they'll
be able to sort out on their own without too much of a delay. The
good news for AMD is that if it's *just* a performance related issue
then it shouldn't affect platform development at all. Chipset and
motherboard makers should be just fine with the pre-production
samples, albeit with somewhat reduced levels of performance.

-------------
Tony Hill
hilla <underscore> 20 <at> yahoo <dot> ca

On Tue, 21 Feb 2006 16:37:08 -0500, Yousuf Khan <(E-Mail Removed)> wrote:

>George Macdonald wrote:
>> Hmm, I wonder if this why they signed up for the Rambus license?... to get
>> some engineering help?
>>
>
>It's likely whatever problems they were having was based on their own
>designs, so they'll have to fix their own bugs. Rambus might be there
>for future revisions the DDR controller. Hell, it might even be there to
>help out with the ZRAM controller. :-)

What could have been been wrong with a DDR2 controller? Since they'd
already done DDR, I'd guess the PHY - IOW a "cell" which Rambus could pull
off the shelf and AMD could slot in. You have to figure that AMD has umm,
collaborated with the likes of Micron to get it working right first...
before the excrement hit the blades.

--
Rgds, George Macdonald