P4 Vcore

As you can see I just got 2.53GHz to 3GHz (3:4 divider so thats 420MHz ram!). I think it's a good overclock, infact its the highest I can do with 3:4. Problem I have to use 1.7VCore.

I have been reading some storys and maybe someone can tell me if its true or not. Some people say that 1.7 VCORE makes your CPU die a lot faster, like one guys P4 2.0 died in 6 month with using 1.7 VCORE. I have water cooling and my temps are 45 FULL LOAD. Do you think its worth or safe to use 1.7 or not? Let me know what you think, thanks.

At 1.7v with excellent cooling, the risk of CPU failure is fairly low. Of course, overclock at your own risk, there are no guarantees. I would not recommend going higher than 1.7v though. Some of those Asus mobos overvolt a little. Is 1.7v what you have set in the BIOS or is it actual measured vcore?

It's what I set in BIOS, also its the highest I can put. How do I know what its at for sure. Is 45 on FULL good cooling or not?

I think you are safe enough, but as Batboy pointed out there are no guarantees once you start jacking the Vcore up. I also feel 1.7V is the most that is advisable, cooling aside. I run 1.65V on my rigs, but if it made the difference between a good OC and a great OC for my particular equipment I would probably run 1.7V.

A 45C Asus P4 reading is to my mind high for water cooling, but low in absolute terms. When I ran an Asus board in my rig with a 1.6a @ 2.4GHz on 1.65V I read 45C max with a Thermalright AX-478 and a 37cfm 80mm fan. Myself I would like to see lower temps to make the expense and hassle of water cooling seem more worthwhile, but the cpu will be happy enough at 45C. Considering the clock rate and voltage you are running it's indicative of good cooling. And as the risk involved with higher Vcores comes as a result of the electrical rather than thermal aspects of the situation cooling the chip to lower temps in reality doesn't make a huge difference.

larva, you have to take ambient into the picture as well. although 45c does seem high. radiator makes a big difference in my rigs though. when i used a big heatercore, i was hitting 35-36c tops (with ambient of around 70F) with 2.05vcoer under load. but changed it to a BIX and it went higher, even up to 40c at times. but with the BIX it was smaller, and i didn't set it up right, fan blowing instead of sucking. i'm about to test my new water setup this weekend though.

also, is it confirmed that all P4 boards use on-die sensor for temps? if so, why is it always different between boards and manufacturers?

Yes, all the P4's use the on die method. Since Intel chips have had this capability for years it would be silly to implement an external thermister, as it would make the board more expensive to manufacture.

Granted, ambient makes a difference, but this is the least of the problems comparing temps from one board or machine to another. The truth of the matter is that the temperature and voltage monitoring on current boards is a 15 cent solution to a 5 dollar problem. What read 40C on my asus board reads 45C on my abit. You really have to test chips or cooling systems on the same board, in the same case, with the same ambient for meaningful results. But nonetheless 45C in Asus degrees is not particularly impressive for water cooling, although certainly good enough for all but the pursuit of really low numbers.

The reason my water cooling temps are go high is because I have a weak fan on my radiator (I don’t want noise).

larva did you make a mistake in your sig.? How did you get 1.6 to 2.6 at 162FSB? What multiplier is your 1.6 and what ratio did you use? I’m just curios.

I'm not Larva, and I don't play him on TV, but the multiplier of the P4 1.6 is 16. At 162FSB with that multiplier he's pulling 2.6GHz on the proc.

One note about SNDS: It does not appear to be temperature related. Some of the first cases were on systems with phase-change cooling where the processor never even ran hotter than room temperature.




BHD

From what I have read it is not the heat but the voltage that kills them. As for myself I will NEVER take one over true 1.6v again. The high voltage may not kill em right away but it has a great chance to give them what I call:

Incuarable Northwood Cancer

Subject to failure and/or loss of overclocking ability in the future.

I didnt know 'incuarable' was a word. To me it seems my Asus P4T-E (ICS) undervolts... could just be my PSU.