Does the Intel G1 stepping suck for overclocking?

[ghost recon88]

So, a quick rundown of the hardware is an Abit AA8XE mobo with the latest BIOS, a P4 521 (2.8GHz), an Antec TruePower 550 watt PSU, and some PQI-Turbo DDR2-5400. I've also got a crappy heatsink but it does a decent job as under a full load at 3.43GHz, it only reached 42 c. So, I was able to complete a 8M SuperPie run with 1.4125 volts. As soon as I get to 3.5GHz, I can't seem to get it stable without a ton of voltage, as in more than 1.4625 volts. What in the heck? How much is safe to crank through this chip? Should it really take that much to get safely above 3.5GHz? It seems a little crazy that it would require .05 more volts to hit an extra 70MHz.

Edit1: It was supposed to be an "E0" revision, but after using the latest CPU-Z, it said it was a "G1" revision. Anyone know anything about these?

 

[ghost recon88]

Doesn't anyone know about these?

 

[thilanliyan]

usually, the further you go with voltage to increase speed the less it makes a difference (ie. diminishing returns). I may not have worded that correctly.

For example I can get my 1.8GHz 3000+ to 2.5GHz (+700MHz) at 1.45 (stock 1.4 so a +0.05v increase) but to get to 2.7GHz I need about 1.55v so at that point I'm approaching the chip's limit.

Maybe you're hitting the chip's limit?

 

[ghost recon88]

I'm wondering as well. It seems anything over stock voltage makes it clock worse. After reading some documents on the Intel website, it appears they added two more 64-bit instructions, LAHF and SAHF, to the CPU. I wonder how running Windows XP-64bit would run, or if it would help with a better overclock. I guess I'll try and sell this one and get an "E0" revision.