Overclocking

[Brodie337]

Pah! Air cooling!

H50 for the win!

[zefrer]

Pah! Off the shelf liquid coolers. Custom loop!

[timw4mail]

If your heatsink doesn't have heatpipes, it's not acceptable for overclocking.

[itlnstln]

Plus you can't smoke it.

Damn, I nearly hosed my HHKB with Diet Dr. Pepper.

[microsoft windows]

Those losers on OCN weren't helpful at all when I asked them how to overclock my Gateway2000.

[Konrad]

Well, if you don't like overclock.net there are a lot of alternative forums,
(overclockers.com, overclockersclub.com, extremeoverclocking.com, extremesystems.org, [h]ardforum.com, anandtech.com, blazingpc.com, devhardware.com, many others) ... I just have no clue which is the "best" or is "best" in any given area, lol

[chimera15]

Hah, even I know that one.  You feed the hamsters steroids and the wheel goes faster.

Show Image

What's the deal with that wheel? They have an rpm meter on it? lol  Is it motorized or something?

[HaaTa]

What's the deal with that wheel? They have an rpm meter on it? lol  Is it motorized or something?

Definitely motorized, or at least not hamster propelled, as the RPM don't go down at all when the hamster flies off, sorta increase actually...

[Konrad]

Overclocking your hamsters is one approach.
 
Just using better hardware to start with will give better results.

[D-EJ915]

Here's my Core 2 Quad I just tried overclocking...pretty good result imo...800MHz memory too.

http://browse.geekbench.ca/geekbench2/view/299161

I need to put less CPU grease on lol

[Konrad]

That's +65% overclock on a 2.13GHz (133MHzx16) part?
 
Yeah, those CPU temps do look high. I hope that's not just idling on the desktop. (If it is, then I would personally scale back or get a serious cooling upgrade.  Your Vcore looks good to me, but I'm no expert.)
 
[Edit]
 
I have the same Seagate HDD

[D-EJ915]

I'm thinking the high temp is just from my room being hot as **** , when I opened the door they dropped 4 degrees almost immediately lol.  The temp was basically the same when at stock speed too so I dunno.  I ordered some AS5 to put on so I'll see if that helps, if not I'm gonna send back this ****in tuniq cooler, it's such a pain in the ass to install who knows if it's on right.

http://valid.canardpc.com/show_oc.php?id=1427877

[zefrer]

That's +65% overclock on a 2.13GHz (133MHzx16) part?
 
Yeah, those CPU temps do look high. I hope that's not just idling on the desktop. (If it is, then I would personally scale back or get a serious cooling upgrade.  Your Vcore looks good to me, but I'm no expert.)
 
[Edit]
 
I have the same Seagate HDD

Different guy, stock clock is 2.5Ghz on his (1Ghz overclock)

[zefrer]

I'm thinking the high temp is just from my room being hot as **** , when I opened the door they dropped 4 degrees almost immediately lol.  The temp was basically the same when at stock speed too so I dunno.  I ordered some AS5 to put on so I'll see if that helps, if not I'm gonna send back this ****in tuniq cooler, it's such a pain in the ass to install who knows if it's on right.

http://valid.canardpc.com/show_oc.php?id=1427877

Have you ran prime95/super pi etc to test stability? I very much doubt it's stable if it idles at 60C..

[Konrad]

lol, it's only ~50C (48C-54C), not 60C.  Plus it apparently drops by ~4C when the ambient is cooled down (ie, door to the room is opened).
 
But that's still pretty high.  Especially if that's on idle ... run a test (super pi) to see what happens under stress load.  It's not a useful OC if it's unstable (crashes whenever you actually play games or do anything).  But hey, it's your money, go ahead and cook your CPU, it'll give you an excuse to upgrade in a few months.
 
AS5 is a low performer these days.  For the same money you could get Indigo Xtreme or Shin-Etsu, which both perform much better.

[Rajagra]

AS5 is a low performer these days.  For the same money you could get Indigo Xtreme or Shin-Etsu, which both perform much better.

If someone sees big performance differences between pastes, I suspect they are applying them badly. Hell, you can get good results without paste provided the surfaces are properly flat and clean. Paste can do no more than fill the air gaps so there is an absolute limit to how much good they can do. There isn't a limit to how much harm they can do - put in too much and you just create an insulating layer.

My overclock seems perfectly stable - except the system can no longer come out of standby. Very odd!

[zefrer]

If someone sees big performance differences between pastes, I suspect they are applying them badly. Hell, you can get good results without paste provided the surfaces are properly flat and clean. Paste can do no more than fill the air gaps so there is an absolute limit to how much good they can do. There isn't a limit to how much harm they can do - put in too much and you just create an insulating layer.

My overclock seems perfectly stable - except the system can no longer come out of standby. Very odd!

Well yeah the difference is not huge but there is one. There is actually at least a 4C difference (hardocp.com) from the average to one of the best, in overclocked conditions.

And no you can't go without. That's like going without a condom.

[Konrad]

A lot of research and big money goes into engineering better TIMs.
True, for consumer PCs the differences aren't that great, although they have slightly greater significance when pushing hot overclocks.
 
Cheap TIM pastes aren't very efficient. Save money and get poor results.
 
Pay a little more for better ("premium" or silver-bearing) TIMs like AS5 to get good results, though there's no need to buy into the expensive boutique brands.
 
Expensive performance TIMs are better still, though you'll see ever-diminishing gains.
 
It's about bang for the buck, right? The best TIM in the world won't give any real advantage over a cheap one if it's efficienctly conducting heat to a heatsink that's too small for the job. Still, if you're already going to pay for AS5 then why not pay the same for Indigo X?
http://skinneelabs.com/tim2010part1.html?page=1

[Rajagra]

Interesting link. Last time I built a system TIM pads were mostly useless. I see he didn't lap the CPU. I won't criticise that, as I've never gone that far. But I wonder what results you'd get by lapping the CPU and heatsink really well, maybe using DMT diamond plates, or even Shapton stones. They are very, very flat (sandpaper on glass just can't compare) and go to insanely fine levels of abrasive, e.g. 1/2 micron. You wouldn't buy these just to lap one install, but someone who was building performance systems regularly just might get some.

The finest diamond plate I have is 3 micron/8000 grit. I may have to play with an old system. :madgrin:

[D-EJ915]

FWIW it went up to around 60-4 C if I remember when doing hyperpi or whichever is the newest on all 4 cores, I don't really remember lol.

[Konrad]

Haha, I've polished my HS and IHS to mirror shine, then applied IndigoX with a razor blade.  I seriously nerded out on it here a few months back.
 
And I'm a noob compared to some of those guys.  They remove the IHS entirely (with a blowtorch, lol the only way) for direct-die liquid cooling.

[Rajagra]

Funny you got info from a laser forum, it reminds me of this article that gives a convincing argument against over-polishing the surfaces.

if we made the mating surface of the aluminum block too smooth, we lost cooling efficiency – similarly if it was too coarse. To the extent that if we polished the surface, we got significantly worse performance than when the heatsink surface was matte after lapping with, say, 400-600 grade carborundum.

It's a good reminder that sometimes what seems obviously true ... is not.