Overclocking the T-bird 800

[pillage2001]

I'm at 900 Mhz now with 1.75 V as the core voltage and running at 45 celcius under heavy load. The question is, before I make mroe drastic movement, should I aim for 1Ghz straightaway or do some burn in for a few days? My initial state at 900Mhz was not stable at all but now it's rock stable after stressing it for days. To o/c or not to o/c?? What are the odds anyway..hehe..I've got a FOp32 with thermal grease on it.

 

[tigger80]

don't believe in the burn in hype, but its worth a try. You should try 1.85 volts

 

[pillage2001]

NOpes, no go even at 1.85V at 950 Mhz. ;(

Anything I can do? I can post up till windows but when I start unzipping a file, it'll just hang.

 

[Technonut]

I currently have 3 800 T-Birds with green cores unlocked with the rear window defogger paint. 2 are week 30's, and one is a week 35. All testing was done with a KT7, 128MB Mosel Vitelic PC133, and Alpha 6035 with a Delta 38CFM fan. The temperatures on all of them at 1.85v ranged from a low of 34 Cel to no higher than 40 under load confirmed by an external temperature probe with the tip placed against the core.

Week 30.....Stable Top End...982MHz
Week 30.....Stable Top End...972MHz
Week 35.....Stable Top End...953MHz

The above MHz was reported by running WCPUID

I could run at a gig with the 2 week 30's, but would crash shortly after Windows loaded. I played with all possible multiplier/FSB/RAM adjustments.

You could try a combination multiplier/FSB overclock. Set your multiplier at 850-900, and slowly bump your FSB a little at a time, booting into windows and running a couple of benchmarks each time until you reach stable top end. My system performs very well doing it that way.

EDIT: I saw no additional benefit concerning higher overclocking by burn in.

 

[pillage2001]

I think I'll try the FSB thingy since I can't squeeze out 950Mhz on my PC.