• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Underclocking!

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

electromagnetic

Member
Joined
Nov 21, 2003
Location
Denver, CO
Who's with me?

Just kidding, but I do have my XP underclocked for the time being. It's hot and my computer room has no AC, my machine was just powering off, literally. This is the only thing that seems to fix it. I checked and the temps were within range, about 47C underclocked, 50C normal, I know that's hot but I'm using air and the room gets to about 80-90F on really hot days.

I have 3 fans in the front of my case and 2 in the back. I get quite a bit of dust, and 2 of the front fans are at 7 volts. If I changed those to 12V and assume I would get more of a positive case pressure, would this help? I volted them down because they were kinda loud. I don't care about loud though if I'm gonna fry my CPU.

The motherboard isn't set to protect the CPU, and it's been overclocked to around 2.3 GHz before, so I don't think it's a dud. Maybe it's my power supply? What do you guys think?
 
I'm not really sure what you're asking, but turning those intakes up to 12V would lower your temps. What kind of PSU do you have? How many fans does it run?
 
Just one PSU fan, it's an Antec PSU that came with the case. I think a 300 watt. Is there a big difference in temperature between negative and positive case pressure? That's my question, and to get any other suggestions.
 
It varies for every case. I find that in my Lian-Li PC-61 negative pressure yields lower CPU temps. Try out both configurations in your case, but remember that fans in the bottom and/or front of the case should be intake, fans at the top and/or back of the case should be exhaust (and intake fan in extra expansion slots is okay, though), and that positive pressure will keep more dust out of your case.
 
Using the same chip as you, I've found that it's possible to stably run overclocked, but undervolted.
It's sorta the best of both worlds.

On my NF7-S I'm currently running 220x10 at 1.55v ( default is 1.65v).
At these settings it will Fold or run Prime for 12 hours with no errors or reboots.
 
clocker2 said:
Using the same chip as you, I've found that it's possible to stably run overclocked, but undervolted.
It's sorta the best of both worlds.

On my NF7-S I'm currently running 220x10 at 1.55v ( default is 1.65v).
At these settings it will Fold or run Prime for 12 hours with no errors or reboots.

Thats what I do 80% of the time. :) I run mine 2-2.1GHz @1.35v. Right now I'm running 250x8 and 34 degrees under load in 27 degree room (sooo hot in here :p )
 
Running my 2500+ at stock right now because of the heat here. It was 30C inside the house on the weekend. Kept getting error beeps while playing AvP2 while I had it set to [email protected] and kept losing connection to servers. Guess I'm stuck at stock for a few days until the temps settle down a bit.

Is it just the thoroughbred B's and the mobiles that run better under volted? Or do the bartons do well undervolted as well?
 
I WAS with you. A while ago...like over a year ago I ran my Athlon XP Palomino 2100+ at 2150+ speeds with the vcore at 1.65. Stock was 1.75.

Also, I recently tried running my 1700 DLT3C Tbred B 0310 XPMW at 700Mhz and 1.0 Vcore to see if I could run passively. It worked.

Underclocking+undervolting can be a challange just like overclocking. You still find the maximum clockspeed for the voltage.
 
I guess voltage to the CPU is the biggest source of heat then? I didn't even sit down at my computer last night, went straight for the TV. Tonight though, I'll put my fans back to 12V and lower the voltage to my CPU and see if that fixes my problem.
 
electromagnetic said:
I guess voltage to the CPU is the biggest source of heat then? .
Well, I don't know if it's the biggest, but it is definitely a significant cause of extra heat.
I found that overvolting, even by little steps, caused what seemed to be a disproportionate rise in temps.
I also decided that , fun as it was intellectually, I really wasn't seeing any practical advantage from OCing above a certain level. Although benchmarks would improve, right about 2.2 GHz seemed to be the sweet spot for performance balanced against temp so that's where I decided to stay. I just kept dropping my vCore till it wouldn't run stable ( below 1.5v in my case) and then upped it a bit ( 1.55v) till it would.
So far, that approach seems to be working just fine...with a room ambient of around 25-28c my load temps are about 15-20C higher ( this is with one 92mm fan at 7v, pulling through my radiator...no other case fans at all).
 
Ahhhhhhhh, the art of undervolting......I shall try this, could be most helpful in the summer heat.
 
Back