• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Dual cored Opterons! 4cpu's for the price of 2, in 2 !

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

@md0Cer

Senior Member
Joined
Nov 6, 2003
Location
Denver, CO
http://www.theinquirer.net/?article=15605

Here they mentioned dual cored Opterons. Some of you may want to take a look. Now that is how to do hyper threading :D :D

Unfortunatly, that article is not very specific and bassically they just mention it, but it is nice to know they at least have it in the works or are thinking of it.


EDIT:

AMD CEO Hector Ruiz


One of the most powerful things next year is going to be our dual-core product. To me, that's going to really shock the hell out of everyone, because it's going to be hardware-compatible, infrastructure-compatible, pin-compatible. I mean, people that have a 2-P system can slap in a dual-core product and end up with a 4-P system for the price of a 2-P. That's been the biggest drawback, everyone tells me. What keeps them from going from a 2-P to a 4-P system? It's price.
 
Last edited:
It appears that AMD planned to go dual core all along with the Opteron since the current chips have the dual core connections aready in place....:D

Talk about a nail in the Itanic coffin....
 
Hmm...that would explain the high pin count, and if they are going to be compatible with current opteron motherboards, they planned it all along. I have a feeling Intel is going to do the same thing. Isn't Socket-T a 700 something pin socket? My guess is that Socket-T Intel's will have the connections for a second core already in place. Quite interesting how something is dormant in a processor while the develope. That shows how much they think ahead. If you think of it, it is rumored that Northwoods all had Hyper Threading dormant from the start, and Prescott's have dormant 64bit instructions. Now AMD already has the connections in place for a dual cored chip.

Back to the topic of dual cores, it could be slightly difficult to have 2 cores on the current CPU. If you have seen the 1mb cache AMD k8's, you know what I am talking about, they have large cores, with a second core, that would take up just about all of the room on the CPU, dont they need some room for connections inside the PCB of the chip?
 
Stacking. Clever.

'bout time they realize dividing and conquering is how to speed things up. Once you cap FIFO... its the only way to go. AIAO (All in, all out). Wouldn't that be neat.

-Frank
 
but then u have the problem of twice the heat on the same hs.... and double voltage through msfets or do they have ways around all of this??
 
a c i d.f l y said:
Stacking. Clever.

'bout time they realize dividing and conquering is how to speed things up. Once you cap FIFO... its the only way to go. AIAO (All in, all out). Wouldn't that be neat.

-Frank

Stacking! I never though of stacking one over the other. Wouldnt it be harder to control the heat though? You would have twice the heatoutput in that same little core. I think doing a side by side type of thing would be best, more surface area to spread the heat out.

From all the news I have read on the idea of dual cored Opterons is they will be putting them together with the 90nm process which will save a bit of space for em.

Good point! That definatly should save some space. And hopefully reduce the heat output. If you think of it, I can handle about 115 thermal watts out of my 1700 with a crappy Volcano 9 heatsink. If they can get the cores combined to equal no more than 160, with good cooling it is definatly possible.

Perhaps this is what those insanely huge pre-production copper heatsinks were for. I am not sure if you guys remember, but these things were so huge a normal A64 could be passively cooled. That was a 3 or 4 months before production though. If they put a 120mm fan on these huge heatsinks, one with 85CFM or higher, and the heatsinks have good airflow abilities that do not restrict the fan all that much, it definatly ought to be able to handle a good 160 thermal watts or so.

but then u have the problem of twice the heat on the same hs.... and double voltage through msfets or do they have ways around all of this??

My comments on heat are above.

As for the voltage, I guess it would not be that hard, just maybe better mosfets? Again, if I can overclock my 1700 to the point where it consumes enough power for 115 thermal watts (and that is not even maxed out!) , the normal mosfets could probably handle it, but I am sure they might just double the number or modify them?

Nice idea's and questions guys!

Cheers!
-0cer
:cool:
 
heat shouldn't be much of an issue, AMD is already producing and selling 35 watt opterons on their 130nm SOI, the dual cores will certainly be 90nm so it should be even less.
 
I thought they did since they are already using SOI tech in the A64s.

Edit- On the AMD site they state the k8 uses SOI. If I remember correctly AMD spent quite a large sum of money with IBM to develop it and get it working.
 
Oni said:
Does AMD have SOI working? I didn't think they did yet.

All AMD K8's have SOI. That is Opteron, Athlon 64, Athlon 64 mobile, and Athlon FX.

I beleive the origional purpose of SOI is for AMD to get 90nm with less powerloss. The Prescott's strained silicon has some troubles with this, which is why it runs so hot, but supposedly should help once they get to a lower gate length. I beleive after 90nm normal silicon like the P4 northwoods, AMD k7's, etc will not give a gain in stability with less power consumption, which is why AMD and Intel are using alternative methods. I heard a while ago that AMD is working with IBM to get to 65nm and 45nm. I heard about them using something called nickel silicide or something of the sort. That is not confirmed, just a rumor I might add.
 
ahhh, this is what my next upgrade will be. When there are dual cores, many years later, thats when i'll upgrade again :p
 
Back