Saturday, May 12th 2012

TIM is Behind Ivy Bridge Temperatures After All

It's proven: the thermal interface material (TIM) used by Intel, inside the integrated heatspreader (IHS) of its Core "Ivy Bridge" processors are behind its higher than expected load temperatures. This assertion was first made in late-April by an Overclockers.com report, and was recently put to test by Japanese tech portal PC Watch, in which an investigator carefully removed the IHS of a Core i7-3770K processor, removed the included TIM and binding grease, and replaced them with a pair of aftermarket performance TIMs, such as OCZ Freeze and Coolaboratory Liquid Pro.

PC Watch findings show that swapping the TIM, if done right, can shave stock clock (3.5 GHz, Auto voltage) temperatures by as much as 18% (lowest temperatures by the Coolaboratory TIM), and 4.00 GHz @ 1.2V temperatures by as much as 23% (again, lowest temperatures on the Coolaboratory TIM). The change in TIM was also change the overclockability of the chip, which was then able to sustain higher core voltages to facilitate higher core clock speeds. The report concluded that Intel's decision to use thermal paste inside the IHS of its Ivy Bridge chips, instead of fluxless solder, poses a very real impact on temperatures and overclockability.

Source: PC Watch
Add your own comment

219 Comments on TIM is Behind Ivy Bridge Temperatures After All

#1
OneCool
by: cadaveca
You cannot use retention bracket with IHS removed, which makes it hard to give CPU correct pressure
that can be addressed without much problem ;)
Posted on Reply
#2
cadaveca
My name is Dave
by: OneCool
that can be addressed without much problem ;)
You bet. I'm actually kind of hoping though, that we see shims for sale again, to make it much easier for average users to take advantage of.
Posted on Reply
#3
Xzibit
by: erocker
But if you have a 2600k/2700k, there's no reason to get Ivy Bridge.. Unless your current SB chip is a complete dud.
Lower power usage overall. S-B doesnt scale well in power usage once overclocked compare to I-B
Improved Memory controller
PCI-E 3.0 - If you want this then your getting a new board and USB 3.0 native support and other improved perks like improved memory speed support on boards.

People upgrade for farless especially the so-called Enthusiast. 3770K and 2600k/2700k are all mainstream chips. I agree if you have a 2nd Gen i-Chip there isnt much to it. Although if your coming from a 1st gen i-Chip or lower the $10-$20 dollar differance is a no-brainer to go with Ivy-Bridge.
Posted on Reply
#4
marcthpro
Well this mean buyer will have to change the Thermal paste i hope some guide soon tell the buyer how to do it safely :P 4.6Ghz at 1.2V for +20$ then Sandy look Tempting for new buyer
Posted on Reply
#5
Dippyskoodlez
by: nvidiaintelftw
you must be high. We don't need a video for this. The overclock is not lazy at all from 3.5ghz to 4.6, and maybe you should learn another language so you can read these things :p
4.6 isn't that hefty.

However, its a good reference point at displaying differences between a single variable and stock.
Posted on Reply
#6
T4C Fantasy
CPU & GPU DB Maintainer
by: erocker
But if you have a 2600k/2700k, there's no reason to get Ivy Bridge.. Unless your current SB chip is a complete dud.
the thing is if you dont already have sandy bridge go for ivy if you have sandy stick with it.... ivy bridge is close to the same price
Posted on Reply
#7
LAN_deRf_HA
By the time they fix this you might as well get haswell.
Posted on Reply
#8
Xzibit
by: LAN_deRf_HA
By the time they fix this you might as well get haswell.
Seeing people systems specs and reading the forums I havent read many post that this will actually effect them.

Its more of a look we found an issue so we pile on and everyone seams to jump on the bandwagon to justify there purchase or lack of it.
More like a child saying no I want same temps and same volts on I-B that I had on S-B, no no no..

I dont know, I bet 1% of people here it will make a differance to. The overcloakers that arent penny pinching. Could probably count them on one hand but then again they arent buying mainstream chips either.

There goes that theory eh? :banghead:
Posted on Reply
#9
erocker
by: Xzibit
Lower power usage overall. S-B doesnt scale well in power usage once overclocked compare to I-B
Improved Memory controller
PCI-E 3.0 - If you want this then your getting a new board and USB 3.0 native support and other improved perks like improved memory speed support on boards.

People upgrade for farless especially the so-called Enthusiast. 3770K and 2600k/2700k are all mainstream chips. I agree if you have a 2nd Gen i-Chip there isnt much to it. Although if your coming from a 1st gen i-Chip or lower the $10-$20 dollar differance is a no-brainer to go with Ivy-Bridge.
Yes, I know about IB. PCI-E 3.0 doesn't make a difference, USB 3.0 isn't a part of IB but the chipset. IB takes more voltage to overclock and scales about the same.
Posted on Reply
#10
sergionography
I dont think intel would shoot themselves in the arm land do that for no reason. They just dont wanna compete with themselves. Meaning ivy bridge being equal tosandy bridge is a win for them as ivy have a tiny 160mm2 die sizes so their margin is even bigger, what ivy is targeted at is mobile, if u comparesb mobile and ivy bridge mobile u willl notice that ivy achieves much higher frequencies and not ivy ultrabooks will perform like sandy bridge notebooks which is a big improvement.
So ivy takes the mobile market as thats what tocks are practical for as being improved process and lower power/TDP while they keep the high end with sb extreme where for thst market po wer consumption isnt as important as max overclock and oc headroom.
Posted on Reply
#13
Xzibit
by: erocker
I'm not really concerned with power usage since SB really doesn't use that much to begin with.
17 at idle and 69 at load differance isnt exactly same

Well to say your not concern with it is one thing but saying its wasnt there is ignoring a feature and an improvement thats there.

:toast:
Posted on Reply
#14
cadaveca
My name is Dave
by: erocker
IB takes more voltage to overclock and scales about the same.
SB does 4.5 GHz @ 1.2V on avg???
Posted on Reply
#15
Hayder_Master
i just want to know how INTEL with the best engineers and testers miss something stupid like that?????
Posted on Reply
#16
NinkobEi
by: Hayder_Master
i just want to know how INTEL with the best engineers and testers miss something stupid like that?????
Maybe it was intentional? Tired of OC'ers getting stupidly high performance from their CPUs. There's no real-world usage reason to upgrade when all you have to do is plop up the clocks a few Ghz.

If you think about it, today's IB i5's at stock clock are probably comparable to a 4600mhz qx9650. So a 5ghz 2700k is probably comparable to the next gen's processors at stock clock. Basically a free 3 years of RnD.. surely that costs intel some money?
Posted on Reply
#17
qubit
Overclocked quantum bit
by: Hayder_Master
i just want to know how INTEL with the best engineers and testers miss something stupid like that?????
by: NinkobEi
Maybe it was intentional? Tired of OC'ers getting stupidly high performance from their CPUs. There's no real-world usage reason to upgrade when all you have to do is plop up the clocks a few Ghz.

If you think about it, today's IB i5's at stock clock are probably comparable to a 4600mhz qx9650. So a 5ghz 2700k is probably comparable to the next gen's processors at stock clock. Basically a free 3 years of rnd.. surely that costs intel some money?
It had to be intentional and I believe I may have nailed it with the marketers and beancounters: :)

by: qubit
I can't believe Intel spent a decade and poured all those billions into developing their impressive 3D transistors, only to squander that advantage over some crappy TIM! :nutkick:

I promise you it wasn't the engineers that wanted it, but the marketers and beancounters within the company. The arguments must have been something else. Shame none of that leaked out.

I'm so glad I didn't jump into IB and got my lovely 2700K. :D

Right, you heard this prediction here first: Intel will backpedal on this decision now that this is out and make a new version of IB with either better TIM or a proper solder connection. They will then offer an optional recall of these processors soon after for the fixed versions.
Posted on Reply
#18
ensabrenoir
Yup...... Intels main competitor is itself.....overclocked last generation equal current gen.....the true ivy bridge would go toe to toe with has...a...well...full of profit for intel:laugh:
Posted on Reply
#19
PopcornMachine
by: Hayder_Master
i just want to know how INTEL with the best engineers and testers miss something stupid like that?????
I've been assured that they did this on purpose and it was an excellent business decision.... :eek:





I don't agree with them however. I think it was really stupid. :p
Posted on Reply
#20
ensabrenoir
Unless intel has a 'special' ...black/ultra/XX version around the corner without any tatical errors
Posted on Reply
#21
swaaye
It is a mystery why they switched back to thermal paste. It must be cheaper, or easier to assemble, or there's something special to 22nm?

AMD still uses paste, I think.

To me, Ivy Bridge is most interesting in low power applications. Can't wait to see the new ultrabooks. For desktop, bring on Haswell...
Posted on Reply
#22
Jurassic1024
Hahaha. Once I saw and posted the delidding here, I knew it wasn't a power density issue. Intel is too smart for that. Now the question is why use TIM vs Fluxless? If they used fluxless, would 5GHz be too easily obtainable maybe? I don't know, but I can't wait to find out.
Posted on Reply
#23
Jurassic1024
by: cadaveca
You bet. I'm actually kind of hoping though, that we see shims for sale again, to make it much easier for average users to take advantage of.
Won't happen. Two chips alone (3770K and 3750K) are not enough to justify the expense for the R&D, especially with maaaaybe a chance of a new revision. By then, Haswell will be out or close to it.
Posted on Reply
#24
cadaveca
My name is Dave
by: Jurassic1024
Won't happen. Two chips alone (3770K and 3750K) are not enough to justify the expense for the R&D, especially with maaaaybe a chance of a new revision. By then, Haswell will be out or close to it.
Bullshit. A shim is a bit of metal that will brace the CPU agains the heatsink, and prevent damage to the core. NO R&D involved, jsut a couple of measurements, and cutting some metal, or even plastic.

like this:








You can pretty much guarantee that extreme guys are gonna want one.
Posted on Reply
#25
qubit
Overclocked quantum bit
by: cadaveca
Bullshit. A shim is a bit of metal that will brace the CPU agains the heatsink, and prevent damage to the core. NO R&D involved, jsut a couple of measurements, and cutting some metal, or even plastic.

like this:

http://www.overclockersonline.net/images/articles/coppershims/durontop.jpg



http://www.tweaknews.net/reviews/tsr/img/shim10.jpg


You can pretty much guarantee that extreme guys are gonna want one.
That's true, it's a quick thing to design.

Your AMD pic reminds me of when I tried buying a shim for my Athlon XP. I put it on, put the paste on and then the heatsink... and the CPU nearly burned up through lack of proper contact. That particular bit of metal found itself in the bin in no time flat.
Posted on Reply
Add your own comment