What a difference a proper heatsink does

Oh man! Ad-blocking software has been detected! :'(

This website is run by the community, for the community... and it needs advertisements in order to keep running. Blocking our ads means your killing our stats!
Please disable your ad-block, or become a premium member to hide all advertisements and this notice.

But then again, we already knew this

Been waiting nearly two months for a local shop to ship me a socket 1366 adapter for my ThermalRight 120 Ultra eXtreme radiator. Been using the stock intel one for my Core i7 975 Exteme Edition.

Finally got tired of waiting and bought it on frozencpu.com, took just 4 days to arrive from the US, and it wasn't even by Fedex!

Began by removing the stock Intel radiator, and what a poor radiator it is... The contact area on the radiator itself is smaller than the CPU surface, by a fair bit. The thermal paste residue left a circular mark on the CPU surface that was nearly a third of an inch smaller diagonally than the distance between the two edges of the CPU surface!!!

I recall using Prime95 to do some stress tests, and the CPU cores didn't take long to hit 100 degrees at which point thermal throttling settled in. Even turning the PC on, the CPU sat at around 55 degrees per core after a few minutes.

The T.R.U.E. however has a contact surface just slightly larger than the CPU, which after I had slathered it with a thin veil of thermal paste, fit snugly in place.

First things I noted were that the idle CPU temperatires once Windows had loaded, were as low as 40 degrees on two of 4 cores! And when I pumped up Prime95 to go through a full lowdown burn-in test, temperatues barely hit 80 degrees!!!

Holy crap

Now I can finally go ahead and OC this thing. I really don't understand why Intel would make their own signature heatsink have a surface area smaller than the CPU itself... Nevermind the fact that the fan on that thing makes considerably more noise than the near-silent Scythe I have installed on the T.R.U.E.

 

Oh man! Ad-blocking software has been detected! :'(

This website is run by the community, for the community... and it needs advertisements in order to keep running. Blocking our ads means your killing our stats!
Please disable your ad-block, or become a premium member to hide all advertisements and this notice.

Yes it does make a difference, it also make a difference if you fit the properly and apply the paste spot on too.

I have a tuniq tower and a E8400 overclocked to 4GHz after testing with prime95 for 7 hours my max temp was 57 degrees. I had a bit of trouble installing the tower due to the heat sinks on my asus rampage formula mobo and there was a miniscule gap at one side of the heatsink and cpu.

So 3 weeks ago I tried to get it properly installed and reapply new paste which I did. I have cut 10 dgrees of my max cpu temp whilst running prime 48 degrees after 7 hours testing my idle temp of 37 has always remained the same.

good luck with the overclock

 

lol

 

Yeah, I'll give it a go for 4Ghz, I'll be "comfortable" with that speed ;)

Seems like the numbers I have, are pretty close to what other people get with their D0 step i7's as well, and I have the disavantage of having a warm room (28 degrees) and my two GTX295's blowing VERY warm air into the case... (they're single PCB editions, so the fan in the middle blows air across the two GPU's, half the air flowing out from the back of the PC, and the other half into the case...).

I think I need to figure out something for the hot air flowing into the case. Maybe a custom built funnel which directs this air straight out of the case.

Now to decide if I'll use the unlocked multiplier to get my 4Ghz speeds, or if I should begin by pushing the frequency instead. Ahhh, decisions decisions.

 

which version of the D0 do you have? and how much ram and what speed is it?

 

What did I say... A simple paper funnel and two cores are down to 39-40 degrees, and two others down to 42 (roughly a 2 degree drop across the board).

 

Greenbrucelee: I've got 12 gigs of ram, 2gig modules each. CAS latencies are 6-7-6-18.

It says S-spec: SLBEQ on the box. Might be the D0 stepping?

hmmm... I need to look up how to interpret this information once again (last time I bothered, was probably 5 years ago). Made in Costa Rica, packed 29th of May. Version number E63336-001.

Then there's a few more numbers here and there for Prod code, MM#, FPO/Batch...

*edit* oooh, just saw this nice chart... too bad they don't have the i7 on it: http://users.erols.com/chare/cpu_id.htm

 

this is your cpu http://processorfinder.intel.com/details.aspx?sSpec=SLBEQ

I got the same cpu to 4.25 a few weeks ago for client.

core speed was 4276.4
multiplier was 32
bus speed was 133.6
qpi 3207.3

this was with a core voltage of 1.384

under full load after 7 hours of prime were 91, 88, 88, 82 with the TJ max being 9,12,12,18. I added a better cpu cooler which was the thermal right 120 and extra fan inside the antec 1200 and reduced the temps to 85 83 83 77

customer has had no issues at all

 

Yeah, I dropped on that page too, but I thought you might have been looking for some other code off the box.

Currently, the 975's all have the code SLBEQ, which may change in a 115x socket or if they bring the core size down from 45nm...

 

Can't you just tell we work on the software side of life.