How is the memory controller for the x2's?

J

jalbs

n00b
Joined
Apr 8, 2005
Messages
61
I hear the San Diego core can handle 4x512 but wasn't sure if I will need to upgrade to 2x1024 in order to get some OC out of them. Even if the X2 is 2 venice cores is there any chance they fixed the memory controllers on them? What about Toledo vs Manchester?

Thanks in advance.
 
the same memory controller for all s939 revE's. the only difference is the number of cores and cache size :cool:
 
jalbs said:
I hear the San Diego core can handle 4x512 but wasn't sure if I will need to upgrade to 2x1024 in order to get some OC out of them. Even if the X2 is 2 venice cores is there any chance they fixed the memory controllers on them? What about Toledo vs Manchester?

Thanks in advance.
Click to expand...

note that they can "handle" 4x double sided dimms, but you probably won't be able to OC them past DDR400 anyway. And you'll probably have to run 2T. The thing about the old ones was that they could only run at DDR333.
 
Russ said:
note that they can "handle" 4x double sided dimms, but you probably won't be able to OC them past DDR400 anyway. And you'll probably have to run 2T. The thing about the old ones was that they could only run at DDR333.
Click to expand...

I decided to test this overclocking with 4sticks out with my X2 3800+ E6, week 29 on a DFI NF4 Ultra-D board with the 704-bt BIOS loaded. I hit 275 mhz 1:1 at 2t with 4 x OCZ TCCD sticks at 2.5-4-3-7. This is SuperPi 32m stable.

275mhz at 2t with 4 sticks at 2.5-4-3-7 is not bad. Could be the new CPU's are whole lot better or maybe I just gave some really good RAM. :cool:
 
i cant OC my memory near as well with my X2 as i could with my San Diego. I would say the Memory Controller is somewhat effected by the fact that its dual core and it is limiting some people's memory overclock. My Ballistix 2X1GB ran perfect 260X10 3-3-3-8 1T with my San Diego, but for the X2 i have to use the 183 divider to get SuperPi stability :*(
 
could just be a difference in the memory controller. it's not unheard of, akin to the core overclocking differently, to have two cpu's with memory controllers that do not perform the same ;)
 
hrm, well i ran Prime95 Blend test overnight and it passed with ram at 3-4-4-10 1T at 2.9Vdimm 1:1 250X10 (Neo 2 overvolts to 2.9 when put at 2.85). Although considering these Ballistix stock volts are 2.8Vdimm at 250Mhz i backed it down this morning only to see it fail Prime Blend immediately at 2.8Vdimm wtf? I put it back up to 2.9Vdimm and it passes. Could the memory controller be causing this kind of problem or is it the ram? Should i consider an RMA to Crucial cuz it seems a bit fishy to have to overvolt the ram to get it to run in spec?
 
You must log in or register to reply here.
Back
Top