"ASUS P5Q Deluxe discussion".

[platojones]

Thanks for that. I wonder if the same applies to the e8000 series dual core processors on the P45 chipset? I also wonder that if it only applies to the quads, why that would be, and if that means that the Vtt circuitry of the quads are more robust than the dual cores?

**EDIT - I just DL'd the latest version of the e8000 data sheet, and it also now says 1.2v standard, 1.14v min and 1.26v max for Vtt. I wonder why it's different for the 4 series chipsets? Well at least I now know up to 1.26v is safe, I was using 1.25v sometimes during benches and was slightly concerned I was over the max. Good news indeed.



It is quite concerning that it was initially fine for a period of time, then there was a non shutdown issue, then finally complete instability. I would be quite concerned that the cpu was damaged, as nothing else really explains the rapid degradation to complete instability.

I would have thought Asus would have tested all processors with their bioses, and it is disappointing this situation has arisen. If the cpu is damaged, along with truehighroller's and no doubt countless others globally, Asus will have a lot to answer for, and they should be replacing all affected cpu's immediately, along with immediately removing that option from the BIOS and any future BIOS releases.

No problem. It's good to know what the 'official' limits are when overclocking anyway. Sounds like you are in good shape. From now on, these datasheets will be the first things I read before attempting to OC on a new chip. I was holding back on upping the Vtt because I had read somewhere that it was 1.15v for the Q9000...but when I read the datasheet, I realized that max only applied to the 3 series chipsets...not the 4 series like the PQ5's. So, I took your advice and upped the Vtt 2 notches to 1.18. Since I did that, I have had no failed posts and was able to prime 12 hours while reducing my DIMM volts a full 1v...sweet deal.

As far as the significant difference in max Vtt between the 3 and 4 series...I can only guess, but I suspect that the ICH10 NB was more than a simple die-shrink...I'm totally guessing here, but I think the ICH10 has much tighter control of voltages, so it allows a higher Vtt without risking excessive transient excursions of Vtt...just a guess though...

[EDIT] Nope, you folks are right...I didn't reduce Vdimm by 1v...more like .1v...still pretty good, eh?[/EDIT]