Memory or CPU?

[MikeyLikesItSI]

Hey guys. Just wanted to ask peoples opinion.

My system specs are in my sig. I haven't seen any signs of thermal throttling, and my temps are pretty consistant. I run 242FSB day and night without a hitch. If i bump up to 245 or higher, when i try dual prime stability, she just locks up. I don't get any sort of prime errors, it just freezes within seconds. I've got PQI 256x2 (not the tccd) at 3.0vmem. Dividers don't help at all, nor does relaxing the timings. I'm running 1.55vcore, and i tried a suicide run of 1.65 to try for 250fsb. i have my pci/agp lock on 33/66, and agp volts at 1.7.

So, my question to all you guru's out there (or anybody with experience) is, does this sound like a limitation from memory? or from CPU? I'd like to think its the memory, but using dividers doesn't help. But i thought if it was the CPU, i'd see errors in prime instead of straight lock up.

 

[Helsyeah]

maybe it is a limitation of the vcore circuit on the mobo... from what i have gathered from following various threads on those 3.2e they beat the crap out of any s478 vcore circuits.. this can be partially fixed with cap mods and adding heatsinks to all of your mosfets

 

[MikeyLikesItSI]

All my mosfets are sinked, with a 90 CFM 120mm fan blowing directly on it all. I have the 875P-T (hybrid 875 chipset, lga775 socket) BigStan achieved somewhere around 270fsb while using the unlocked multi option on this mobo. My voltages are right where they should be. of course, unless there was a drop somewhere that mobo monitor doesn't detect.

Appreciate the response, thank you.

 

[Styyn]

Try running memtest86+ with the 1:1 divider. Keep bumping it up 1mhz at time and looping test #5 until you get errors. That should give you a pretty good idea of the ceiling for your ram. From there it's just process of elimination

 

[jenko]

Im no expert but its seems alot of people struggle to get over 240 fsb prime95 stable on air cooling without volt mods.

1.55Vcore is pretty high for a pressie on LGA775 im not sure? think i can do 241fsb on 1.45Vcore on my 3.0e which seems a limits on 478 with the above mods.

Batboy`s doing well on his setup.

 

[Helsyeah]

I have the 875P-T (hybrid 875 chipset, lga775 socket)

doh, guess i should have paid a little more attention before posting lol

 

[MikeyLikesItSI]

ok styyn, i'll try memtest. i suppose thats the hardest mem stressing program (give or take)?

jenko: i'm on H2O, not air.

 

[MikeyLikesItSI]

well. i just memtested the snot out of it. and found that my memory starts producing errors around 230 fsb. amazing too, because i'm dual prime stable at 242. 5/4 Mem divider would take away errors up to 255fsb...... so i guess its safe to say its not the memory, cause i can't get over 242 fsb even on 3/2 divider

 

[neongenesis]

Its never been any prob for me with a little bit stronger cpu

 

[jszent]

Hey guys. Just wanted to ask peoples opinion.

My system specs are in my sig. I haven't seen any signs of thermal throttling, and my temps are pretty consistant. I run 242FSB day and night without a hitch. If i bump up to 245 or higher, when i try dual prime stability, she just locks up. I don't get any sort of prime errors, it just freezes within seconds. I've got PQI 256x2 (not the tccd) at 3.0vmem. Dividers don't help at all, nor does relaxing the timings. I'm running 1.55vcore, and i tried a suicide run of 1.65 to try for 250fsb. i have my pci/agp lock on 33/66, and agp volts at 1.7.

So, my question to all you guru's out there (or anybody with experience) is, does this sound like a limitation from memory? or from CPU? I'd like to think its the memory, but using dividers doesn't help. But i thought if it was the CPU, i'd see errors in prime instead of straight lock up.

Yes I do believe you are at your limits with ddr 3200 1:1. If not then your chip is maxed out,usaully being unstable means that your chip has reached top end...........