soccerballtux
Lifer
- Dec 30, 2004
- 12,553
- 2
- 76
But if majority of people are able to push it over 4 ghz, i should be able to atleast push it about 3.5
yeah, the odds are unlikely your CPU can't do it, it's probably something else going haywire.
But if majority of people are able to push it over 4 ghz, i should be able to atleast push it about 3.5
The max safe temp for that CPU is 62. If you are already hitting 60, you don't want to start messing with voltage until you get better cooling.
My 1045T (same CPU, different binning) didn't want to go much above 3.5Ghz. Silicon Lottery indeed!
On this subject i have read phenom II's specially Thubans are really good OC'ers. And specially with Multiplier. I am gonna work with FSB as i had no success with setting NB speed and voltage and see what happens
THanks man you were really help ful.NB overclocking is basically pointless. Leave the HyperTransport alone.
CPU-NB on the other hand, yes this is worthwhile. Start with just the core frequency though until you get that to the max, then turn off that OC and swap to the CPU-NB
you probably already know this and are referring to the CPU-NB/L3 cache.
definitely start with one variable at a time. That's the only way to narrow it down. If you add 3 variables, you don't know which change it was that causes instability.
Start with multiplier overclocking, IE keep the bus at 200mhz and just bump the multiplier.
To give you an idea, I needed either 1.44v or 1.4725v to hit 4ghz on my Ph2-965. I gave 1.3v IIRC to the CPU-NB to hit 2.6ghz.
You can use PhenomMSRTweaker to do overclocking in software very easily.
I didn't use that to do the CPU-NB overclcoking, I did that exclusively in the BIOS. But being able to change the cores inside windows made overclocking much easier.
I'd hesitate to label any chip as having a safe overclock range that "everybody else" gets.
typically people brag about their success, the failures go unnoticed.
Often people will just run a bench or two and the system isn't really ever stable.
