Holy crap this thing undervolts like a champ

Lean L

Diamond Member
Apr 30, 2009
3,685
0
0
92520708.png


Wow... Keeps it cool as fuck too. I assume that this is fine if it's stable? Undervolting processors won't damage them?
 

Rubycon

Madame President
Aug 10, 2005
17,768
485
126
It would be nice to know what you're undervolting!
A light bulb undervolted will run for a thousand years, stay cooler and not attract as many bugs - too. ;)
 

Ika

Lifer
Mar 22, 2006
14,264
3
81
It would be nice to know what you're undervolting!
A light bulb undervolted will run for a thousand years, stay cooler and not attract as many bugs - too. ;)

CPU-Z says Q9550. 2.85ghz at 0.94v, that's pretty crazy stuff.
 

VirtualLarry

No Lifer
Aug 25, 2001
56,587
10,227
126
Wow, nice. I wonder how much wattage you are burning (or rather, not burning) at that voltage.
 

Rubycon

Madame President
Aug 10, 2005
17,768
485
126
CPU-Z says Q9550. 2.85ghz at 0.94v, that's pretty crazy stuff.

The OP had no image there at first. Probably our network as it's running slower than usual. Probably alien frogs pooping all over the transponders. :D
 

Lean L

Diamond Member
Apr 30, 2009
3,685
0
0
Right. Assuming that this is stable, would it increase the cpu's life?
 

frostedflakes

Diamond Member
Mar 1, 2005
7,925
1
81
I don't really overclock these days, just undervolt at stock speeds. If you don't plan to OC, there's really no reason not to IMO. Great way to cut down on heat and noise.

edit: And yeah, it would probably increase life of the processor, but you're talking about going from a lifetime of 10 years, for example, to 12 years. Even at default voltage these things are designed to last, it would probably be obsolete before it dies. The main benefits of undervolting are lower temps, noise, and power consumption.
 
Last edited:

MagickMan

Diamond Member
Aug 11, 2008
7,460
3
76
Isn't that a Q9550S? Even so, it's mighty impressive.

Edit: Ah, no "ES" at the end of the Spec, nm.
 
Last edited:

Kenmitch

Diamond Member
Oct 10, 1999
8,505
2,250
136
My old 4ghz(1.300v) Q9550 (1.175 VID) would run stock speed around the same voltage as yours. 100% stable but I tested it with more than Prime95 tho. Try something a little more demanding just to make sure :)
 

Rubycon

Madame President
Aug 10, 2005
17,768
485
126
Careful - dropping volts with same clock speed is technically the same as overclocking or speed margining. Silent corruption symptoms can result (same if o/c) if you don't ensure stability!
 
Dec 30, 2004
12,553
2
76
Careful - dropping volts with same clock speed is technically the same as overclocking or speed margining. Silent corruption symptoms can result (same if o/c) if you don't ensure stability!

while technically correct, your post is overrated in the communication of this point. Sometimes I get the feeling the leet overclockers do things like tell the noobs they have to have sub 55C temps to ensure a "safe" overclock to scare the competition away.
 

Rubycon

Madame President
Aug 10, 2005
17,768
485
126
while technically correct, your post is overrated in the communication of this point. Sometimes I get the feeling the leet overclockers do things like tell the noobs they have to have sub 55C temps to ensure a "safe" overclock to scare the competition away.

No, the same rule applies - it should be tested just as if it were overclocked! Plain and simple as that.
 

Sahakiel

Golden Member
Oct 19, 2001
1,746
0
86
well yeah but it sounded a lot worse when you said it the way you did

How does it sound worse? It's exactly the same situation as an overclock. You don't raise the voltage just for fun. It's to try and force signals to stabilize above threshold earlier.
When you drop the voltage, the minimum time period to guarantee a proper read increases. If you drop it too low, you run into the same signal problems as an overclock. The only difference is you're doing it at a cooler temperature.