Sorry I failed to mention, but i'm running different settings now. I tested with orthos for 6 hours and it ran stable. I'm now running at 2:3 ratio with FSB333 and my memory @ 1000MHz. Cpu vcore is @ specs max of 1.325. But I upped the memory to 2.30v and the NB to 1.45v. Ya think maybe i'm getting hooked on this stuff? Oh, ran the memory bandwidth thingy on Sandra and both readings were in the 6600's. Temps under load were in the mid 50's, idle 38c or there abouts.
You probably don't need that much Vcore to run 333FSB. Drop it down a bit and see if it stays stable. I run my E6600 at 334FSB daily with 1.26V
Whooah dude, not true. Electromigration can cause some serious damage to your processor/ram/mb/whatever regardless of temperature. In simple terms, it occurs when the current density is more than the weakest component in the circuit can handle. My advice is to keep your vcore and other voltages within manufacture specs if you care about the longevity of your hardware.Originally Posted by ol'norton
For Intel chips, use the processor finder to find the recommended vcore range. For memory, consult the manufacture of the RAM (the max I have seen for DDR2 is 2.2 V). I dunno where you'd look for vcore's for your board like FSB, NB, SB, and ICH.
Do you know anyone this has happened to? I have yet to here of anyone killing a Core 2 Duo with current alone. And mine has done a lot of hours above 1.5V
Memory slots
Any truth to the belief that moving your 2 sticks of memory to the black slots(farthest away from cpu) will possibly give you greater overclocking ability?
First I've heard of that... try it and see
Read it in more than one forum. Not sure if I want to go through the trouble though. After almost 2 weeks of learning the overclocking wizardry, I think I can do bios settings blindfolded. I'm overclocking both the cpu and ram. Doesn't make sense to me to do one without the other. Runs just fine at FSB300 and Dram at 900. Going to 333 and 999 won't cut it. I'm trying to avoid too much voltage. Plus it seems any combination that equals 2.5 to 1 isn't an option. 2 to 1(1:1) and 3 to 1(2:3) does work, as long as I'm not exceeding the limits of my memory. So, right now i'm trying 320 FSB @ DDR2960. Upped memory v to 2.25 and NB to 1.45. Timings @ 4-4-4-15. So far, so good.
@aceshu - sounds good... the timings typically work best according to this formula:
x-x-x-(3x)-(x+1)
In other words, if x=4, they become: 4-4-4-12-5 and if x=5, they become 5-5-5-15-6 etc.
Also, I see the date on your post is quite a while ago. I'd actually recommend that you drop your memory down to 1:1 which will very likely allow you to drop your NB voltage and likely the mem voltage.
There are two kinds of o/c'ers:
1) Those who want every last MHz out of their equipment at any cost
2) Those who want to run as fast as possible using the minimum amount of voltage
Where did you come up with this "formula"? How did you test to see if it works better than other settings?
I would like to know also where this formula came from?
I have been beta testing ram for over 4 years now and never seen this equation!
@nutcase: One of the folks at Crucial told me that when I was asking about timings for my RAM. I found it to be pretty good in my own systems.
Spammer raised this thread and I wanted to confirm that the spam-o-matic is working.
Now some food for spambots:
User Details:
Username: jan83fry
User ID Number: 29646
User E-mail: jan83fry@hotmail.com
User Registration IP Address: 182.68.220.235
User IP Address for Selected Post: 182.68.220.235
Posting Permissions
Reply With Quote
Bookmarks