Anandtech attempts to sort the mess of AMD core revisions.

http://anandtech.com/cpuchipsets/sho...spx?i=2418&p=3

The Socket 939 situation description is OK, but I thought it was
even more complex than that...

Grumble wrote:
> Anandtech attempts to sort the mess of AMD core revisions.
>
> http://anandtech.com/cpuchipsets/sho...spx?i=2418&p=3
>
> The Socket 939 situation description is OK, but I thought it was
> even more complex than that...

in that report they didn't mention the new opteron sockets (over 1000
pins?) coming out next year did they... ? or maybe they did mention it
and i'm be a blind arse.

On Wed, 18 May 2005 22:30:48 +1200, gimp <(E-Mail Removed)> wrote:

>Grumble wrote:
>> Anandtech attempts to sort the mess of AMD core revisions.
>>
>> http://anandtech.com/cpuchipsets/sho...spx?i=2418&p=3
>>
>> The Socket 939 situation description is OK, but I thought it was
>> even more complex than that...
>
>in that report they didn't mention the new opteron sockets (over 1000
>pins?) coming out next year did they... ? or maybe they did mention it
>and i'm be a blind arse.

The exact nature and release date of those sockets are just rumors at
this point, so no they did not mention them.


Also, FWIW it looks like the 1000+ pin socket rumors are turning out
to be false. The only sockets that have been shown so far have been a
new 939 pin socket for DDR2 memory support on the Athlon64 and a new
small-form-factor 754-pin socket for mobile processors. By all
accounts the Opteron will stick with Socket 940 until at least
mid-2006.

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

Tony Hill wrote:
> Also, FWIW it looks like the 1000+ pin socket rumors are turning out
> to be false. The only sockets that have been shown so far have been a
> new 939 pin socket for DDR2 memory support on the Athlon64 and a new
> small-form-factor 754-pin socket for mobile processors. By all
> accounts the Opteron will stick with Socket 940 until at least
> mid-2006.

eek - only one year left..? but i suppose 1 year is a long time in
socket-years. that's actually what i'd heard before, at about that time
it would change...

ps. what does FWIW stand for..? From What I....

>
> ps. what does FWIW stand for..? From What I....

For What It's Worth.

On Wed, 18 May 2005 09:52:18 +0200, Grumble <(E-Mail Removed)> wrote:

>Anandtech attempts to sort the mess of AMD core revisions.
>
>http://anandtech.com/cpuchipsets/sho...spx?i=2418&p=3
>
>The Socket 939 situation description is OK, but I thought it was
>even more complex than that...

Yes, it is more complicated than that... possibly due to cleaning out of
parts which are binned down... or maybe even "highly overclockable":-)
Then there's the non-boxed stuff which clutters the picture. The current
shipping boxed s939 3500+ showing at Newegg is a Clawhammer core with 512KB
L2 and is not shown in Anand's table, though they hint at "variants". I'm
not sure if this summary http://www.c627627.com/AMD/Athlon64/ is strictly
all correct but it does give a better picture of things... I think.:-)

--
Rgds, George Macdonald

On Thu, 19 May 2005 16:26:39 +1200, gimp <(E-Mail Removed)> wrote:

>Tony Hill wrote:
>> Also, FWIW it looks like the 1000+ pin socket rumors are turning out
>> to be false. The only sockets that have been shown so far have been a
>> new 939 pin socket for DDR2 memory support on the Athlon64 and a new
>> small-form-factor 754-pin socket for mobile processors. By all
>> accounts the Opteron will stick with Socket 940 until at least
>> mid-2006.
>
>eek - only one year left..? but i suppose 1 year is a long time in
>socket-years. that's actually what i'd heard before, at about that time
>it would change...

Socket 940 was first introduced in April of 2003. A virtually
unchanged socket lasting for 3+ years is basically unheard of for
high-end processors.

>ps. what does FWIW stand for..? From What I....

For What It's Worth.

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