[grahamcrackuh]

Alright guys,

I want to have a definitive conversation over the safe zones for Vcore max on a Q9550 and the VTT max for a Q9550.

My current position is that for 24/7 use keep the Vcore at 1.4 or less (because with LLC enabled, it can spike above 1.45, which is where intel says it starts to break down).

My position about the VTT is keep it below 1.4 for 24/7 use before degradation might occur.

What do you guys think about these values? I'm running my VTT at 1.36 right now, its the only way I can hit 4.3 on my 9550 and I was wondering if you'd consider that safe for 24/7 use while hopefully having it last a couple years still.

 

[FSF-Foxhound]

I want to run my chip at 1.55 VCore for months, to see if it really degrades. I think 1.45 is a tad low. If the i7's use the same hi-k process, then the yorkfeilds should be safe up to 1.55, no matter what intel says..

my VTT's @ 1.36 i think..

imo, max 24/7 on proper cooling for the voltages, is 1.55 VCore, and 1.45 VTT.

 

[grahamcrackuh]

Quote:

Originally Posted by FSF-Foxhound I want to run my chip at 1.55 VCore for months, to see if it really degrades. I think 1.45 is a tad low. If the i7's use the same hi-k process, then the yorkfeilds should be safe up to 1.55, no matter what intel says.. my VTT's @ 1.36 i think.. imo, max 24/7 on proper cooling for the voltages, is 1.55 VCore, and 1.45 VTT.

I think cooling is irrelevant, voltages will degrade a chip much faster than slightly high temps. anandtech fried their 9550 during reviewing it @ 1.42 vcore and 1.4 vtt.

 

[FSF-Foxhound]

Quote:

Originally Posted by grahamcrackuh I think cooling is irrelevant, voltages will degrade a chip much faster than slightly high temps. anandtech fried their 9550 during reviewing it @ 1.42 vcore and 1.4 vtt.

C0, C1, or E0?

Also, intel might of sent them a engineering sample, which probably was more feeble.

If cooling is irrelevant, then how come my E8400 lasted 6 months @ 1.55 Vcore?

I need an explanation.

edit: where is this annandtech page, and what motherboard. Because, lower quality motherboards and capacitors can degrade a CPU at higher voltages.

 

[ericeod]

You should be fine with vtt at 1.36v. I wouldnt worry about it at all. Once you start getting to 1.4v is when you have to worry. And I think anyone running 1.4v wouldnt really need that much for any OC (except maybe extreme LN or something).

 

[FSF-Foxhound]

I agree ericeod. Myself, i don't really even need 1.36 VTT. I bet i could put it on stock.

 

[tehmaggot]

I limit myself around 1.39ish on vcore and 1.36 on VTT. Neither are really necessary, but those are the highest I ever ran on boards with horrible VDroop. I'm currently running 4.2Ghz with 1.35 VCore (around 1.32, idle and load), and 1.36 VTT. I'm not sure if the VTT is even necessary, but I'll leave it there for now.

 

[ericeod]

Quote:

Originally Posted by FSF-Foxhound I agree ericeod. Myself, i don't really even need 1.36 VTT. I bet i could put it on stock.

Yeah I ran 1.20v vtt (actual, in bios it was set to 1.26v) at 4.2GHz with my Q9650 and Rampage Formula. But I wasnt pushing as high a FSB as the OP.

Here it is at 1.17v vtt at 4.1GHz:

 

[grahamcrackuh]

@ foxhound and eric

I found that I really didn't need to increase vtt at all until i went for FSB 490 +. So that's probably why you guys are able to keep it stock. The main purpose of VTT is to deal with higher fsbs anyway, since its the termination of binary string voltage that the fsb sees first and last in a set.

 

[supaspoon]

Quote:

Originally Posted by FSF-Foxhound I want to run my chip at 1.55 VCore for months, to see if it really degrades. I think 1.45 is a tad low. If the i7's use the same hi-k process, then the yorkfeilds should be safe up to 1.55, no matter what intel says.. my VTT's @ 1.36 i think.. imo, max 24/7 on proper cooling for the voltages, is 1.55 VCore, and 1.45 VTT.

I'd certainly be interested to see as well. While yeah voltage will still 'damage' without necissarily being helped a long by temps, I certainly doubt cooling is 'irrelevant'.

I'm sure these chips can handles those volts easy enough on a phase, but I think wether it can last up there on the temps from a good water setup is yet to be seen.

 

[ericeod]

Quote:

Originally Posted by grahamcrackuh @ foxhound and eric I found that I really didn't need to increase vtt at all until i went for FSB 490 +. So that's probably why you guys are able to keep it stock. The main purpose of VTT is to deal with higher fsbs anyway, since its the termination of binary string voltage that the fsb sees first and last in a set.

Yeah, thats what I was assuming. I figured you needed higher because you were running a higher FSB (4.3GHz with 505 FSB).

 

[grahamcrackuh]

Quote:

Originally Posted by supaspoon I'd certainly be interested to see as well. While yeah voltage will still 'damage' without necissarily being helped a long by temps, I certainly doubt cooling is 'irrelevant'. I'm sure these chips can handles those volts easy enough on a phase, but I think wether it can last up there on the temps from a good water setup is yet to be seen.

I apologize, I didn't mean cooling is irrelevant, but you will have to have readings of at least 79 degrees I would imagine in coretemp before you have to worry. The intel specified Tcase would be the reading at the very top of the cpu, not tjunction that coretemp gives you. Normally tjunction (since it's inside the chip) will be a decent amount hotter than tcase (probably around 6 - 10 degrees C).

 

[supaspoon]

Quote:

Originally Posted by grahamcrackuh I apologize, I didn't mean cooling is irrelevant, but you will have to have readings of at least 79 degrees I would imagine in coretemp before you have to worry. The intel specified Tcase would be the reading at the very top of the cpu, not tjunction that coretemp gives you. Normally tjunction (since it's inside the chip) will be a decent amount hotter than tcase (probably around 6 - 10 degrees C).

Ahh right on. Yeah I guess I totally misunderstood what you were getting at.

 

[grahamcrackuh]

So anybody wanna throw out some definite opinions on where they think you should keep the 9550 Vcore and VTT at to be safe? (PLL obviously doesn't matter too much, below 1.57 I'm sure most of us would agree, with a CPUreference voltage @ about .63 - .67

 

[FSF-Foxhound]

The VCore should be 1.55, although, to make you guys feel better, 1.45 should be put, because I don't think most people know that the same Hi-k manufacturing on the Q9550s are used in the i7s. Basically, it should be safe upwards 1.55 VCore.

PLL, duno about the q9550s but i ran my E8400 @ 1.8-1.9 PLL for months, no sign of degradation.