Wednesday, April 25th 2012

Ivy Bridge Temperatures Could Be Linked To TIM Inside Integrated Heatspreader: Report

PC enthusiasts with Ivy Bridge engineering samples, and reviewers at large have come to the consensus that Ivy Bridge is a slightly warmer chip than it should be. An investigation by Overclockers.com revealed a possible contributing factor to that. Upon carefully removing the integrated heatspreader (IHS) of an Ivy Bridge Core processor (that nickel-plated copper plate on top of the processor which makes contact with the cooler), the investigator found common thermal paste between the CPU die and the IHS, and along the sides of the die.

In comparison, Intel used flux-less solder to bind the IHS to the die on previous-generation Sandy Bridge Core processors in the LGA1155 package. Attempting to remove IHS off a chip with flux-less solder won't end well, as it could rip the die off the package. On the other hand, the idea behind use of flux-less solder in CPU packages is to improve heat transfer between the die and the IHS. Using thermal paste to do the job results in slightly inferior heat transfer, but removing IHS is safer. One can be sure that making it safe for IHS removal couldn't have been the issue behind switching back to conventional thermal paste, as everything under the IHS isn't user-serviceable anyway, and off limits for them. Perhaps Intel kept extreme overclockers in mind.

Source: Overclockers.com
Add your own comment

97 Comments on Ivy Bridge Temperatures Could Be Linked To TIM Inside Integrated Heatspreader: Report

#1
qubit
Overclocked quantum bit
larsjrg said:
Warmer? My new i5-3550 runs same temperatures with the intel stock cooler at the 2500K do, about ~62-63C on load at stock clock speeds. I'm a happy buyer with a Ivy Bridge CPU using less power than my previous Athlon II x4, which was about ~48C on load with AMD stock cooler...

So what is all the fuss I read about increased temperatures? I certainly do not see the proof when it comes to my Ivy Bridge compared vs. old reviews on Sandy Bridge. (Different intel stock coolers might contribute a few degrees differences)
They get hot when overclocked hard - much hotter than Sandy Bridge, that's the problem.

IB seems to be more optimised for power efficiency at stock settings and great IGP performance, rather than being tuned for the enthusiast overclocker. It's not really surprising, when you see that computing is going mobile more and more and us enthusiasts make up a miniscule percentage of Ivy Bridge's buyers. It's a real shame, but that's life.
Posted on Reply
#2
larsjrg
qubit said:
They get hot when overclocked hard - much hotter than Sandy Bridge, that's the problem.

IB seems to be more optimised for power efficiency at stock settings and great IGP performance, rather than being tuned for the enthusiast overclocker. It's not really surprising, when you see that computing is going mobile more and more and us enthusiasts make up a miniscule percentage of Ivy Bridge's buyers. It's a real shame, but that's life.
Well thanks for that clarification :) Even as a non-overclocker, I do read all of these overclocking articles, just to see where the limits are. It is fun but also confusing at times when you just want some details on the stock performance :confused: as nobody seem to care for it really (at least in the reviews).

Anyhow, as you say Ivy Bridge seems to be targeted at a different segment of the market, I'm one of these people. Usually I find them rare in these kind of forums.
Posted on Reply
#3
AsRock
TPU addict
newtekie1 said:
At this point, I'd say using paste instead of solder saves a little more than half a cent per chip. The solder requires special equipment, and and is more expensive than paste.

Even if we say the savings is only say 1 cent per processor, Intel has shipped 75 Million+ Sandybridge processors, and likely will ship close to that number if not more Ivy Bridge processors. So that is a huge savings. One cent on 75 Million processors is $750,000. That is a nice chunk off the bottom line for a change that won't affect but maybe 10% of users.
Just what i was thinking and even if it was 1/2 cent it's still a load of cash.
Posted on Reply
#4
qubit
Overclocked quantum bit
larsjrg said:
Well thanks for that clarification :) Even as a non-overclocker, I do read all of these overclocking articles, just to see where the limits are. It is fun but also confusing at times when you just want some details on the stock performance :confused: as nobody seem to care for it really (at least in the reviews).

Anyhow, as you say Ivy Bridge seems to be targeted at a different segment of the market, I'm one of these people. Usually I find them rare in these kind of forums.
Yes, overclocking is a bit of an art, to get the most out of a chip, which makes things challenging at times.

Also, as a non-overclocker you potentially have an advantage: you'd think that the K chips are the same as the non-K chips appart from the locked multiplier, wouldn't you. Incredibly, they're not: Intel actually disable the odd feature here and there in the K chips. For SB, their virtualization technology was more limited. I have no idea why they'd do this, because you need more processing power for a virtual environment, which would justify overclocking.

I'd be interested to see if anyone can explain Intel's logic on this one, as it doesn't make any sense to me.
Posted on Reply
#5
speedpc
I setup a new IB 3770K for a friend and before doing the water cooling i put on the stock heatsink that came with it, i thought to myself nice little copper center it should do OK. NOPE !!!! everything at stock the idle temps looking at the bios were 50C !! well the first thing i did was clean everything off and put on the good MX-4 that only brought it down to 47C. well i booted into windows for a good old fashion test run. the first and only test was 3DMarks and when it hit the cpu test it shot up to 75C !!!!and still climbing the alarm was going of in the case so i had to shut it down and this was at STOCK !!! Reminds me of the P4 Prescott days :( SERIOULY They definetly need alot better stock heatsink/fan with these. IMO.
Well now He's on water and for a week or so he will run it at:
Intel Core i7-3770K, 4433 MHz (43 x 103) < Using Gigabyte Easy Tune
Field Value
Sensor Properties
Sensor Type ITE IT8728F (ISA A30h)
GPU Sensor Type Diode, CHiL CHL8266 (NV-Diode, 46h)
Chassis Intrusion Detected No

Temperatures (MUCH BETTER)
Motherboard 40 °C (104 °F)
CPU 9 °C (48 °F)
CPU Package 29 °C (84 °F)
CPU IA Cores 29 °C (84 °F)
CPU GT Cores 25 °C (77 °F)
CPU #1 / Core #1 20 °C (68 °F)
CPU #1 / Core #2 29 °C (84 °F)
CPU #1 / Core #3 22 °C (72 °F)
CPU #1 / Core #4 24 °C (75 °F)
Aux 40 °C (104°F)
I mentioned this in my thread that i have noticed my friends GPU runs hotter on this gigabyte motherboard and not sure why maybe its just the sensor.
So if you buy a Ivy Bridge PLEASE spend a little extra and get a good heatsink and fan IMO
Posted on Reply
#6
qubit
Overclocked quantum bit
speedpc said:
I setup a new IB 3770K for a friend and before doing the water cooling i put on the stock heatsink that came with it, i thought to myself nice little copper center it should do OK. NOPE !!!! everything at stock the idle temps looking at the bios were 50C !! well the first thing i did was clean everything off and put on the good MX-4 that only brought it down to 47C. well i booted into windows for a good old fashion test run. the first and only test was 3DMarks and when it hit the cpu test it shot up to 75C !!!!and still climbing the alarm was going of in the case so i had to shut it down and this was at STOCK !!! Reminds me of the P4 Prescott days :( SERIOULY They definetly need alot better stock heatsink/fan with these. IMO.
That's just scary. Doesn't sound like IB is all that if it runs so hot at stock with the stock cooler, does it? Perhaps Intel should put a CPU health warning with that stock cooler? :laugh:
Posted on Reply
#7
ShiBDiB
speedpc said:
I setup a new IB 3770K for a friend and before doing the water cooling i put on the stock heatsink that came with it, i thought to myself nice little copper center it should do OK. NOPE !!!! everything at stock the idle temps looking at the bios were 50C !! well the first thing i did was clean everything off and put on the good MX-4 that only brought it down to 47C. well i booted into windows for a good old fashion test run. the first and only test was 3DMarks and when it hit the cpu test it shot up to 75C !!!!and still climbing the alarm was going of in the case so i had to shut it down and this was at STOCK
I call bs... no reason any "alarms" should be going off at 75C when the TJ Max is 30 degrees beyond that
Posted on Reply
#8
speedpc
It's a great cpu if it has the right cooling, and in this case i think Intel KNEW they ran HOT and should have shipped a proper heatsink/fan. I am going to call Intel on monday and see if these temps are normal (even if i had to put the stock heatsink/fan back on and make a video for them lol) I personally think it's way to fn hot with everything at stock. trust me i tried alot of different things to try to cool it down at stock but it stayed hot and under load forget it they shot way up.
I think the fan was at 1750RPM and when i disabled the fan control in the bios i think it only went up to 2170RPM not nuch help :(
Posted on Reply
#9
speedpc
No BS If the CPU Warning Temp is set in the bios at 70C then it will go off. I had it set to
60C first but even the slight load set that off. need more proof I'll be glad to make a video :roll:


ShiBDiB said:
I call bs... no reason any "alarms" should be going off at 75C when the TJ Max is 30 degrees beyond that
Posted on Reply
#10
ShiBDiB
speedpc said:
No BS If the CPU Warning Temp is set in the bios at 70C then it will go off. I had it set to
60C first but even the slight load set that off. need more proof I'll be glad to make a video :roll:
Then that's ur own fault. Theirs nothing unsafe about that temp
Posted on Reply
#11
qubit
Overclocked quantum bit
ShiBDiB said:
Then that's ur own fault. Theirs nothing unsafe about that temp
You're very negative, aren't you? Sheesh.

It's not "his fault". :rolleyes: Ivy Bridge has been engineered to use significantly less power than Sandy Bridge: it fits into a 77W power envelope instead of 95W. Therefore, it should run cooler at stock not hotter, simple as that.

Having the CPU idle at 50C and then hit 75C+ on running a standard 3D Mark benchmark points to a fault somewhere. Perhaps the mobo is feeding it too much voltage? Could be a lot of things, but the fact remains that the CPU is running too hot and he's correct to flag it.
Posted on Reply
#12
speedpc
Yes i was just stateing my experience with the Ivy Bridge at stock with the stock heatsink/fan. I would be interested to hear if others have noticed the same temps on there 3770K at stock.To me I THINK it's to hot. If anybody has any suggetions on a different heatsink/fan I'll be more than happy to try it :) I like taking things apart especially when it's not mine lol :respect:

qubit said:
You're very negative, aren't you? Sheesh.

It's not "his fault". :rolleyes: Ivy Bridge has been engineered to use significantly less power than Sandy Bridge: it fits into a 77W power envelope instead of 95W. Therefore, it should run cooler at stock not hotter, simple as that.

Having the CPU idle at 50C and then hit 75C+ on running a standard 3D Mark benchmark points to a fault somewhere. Perhaps the mobo is feeding it too much voltage? Could be a lot of things, but the fact remains that the CPU is running too hot and he's correct to flag it.
Posted on Reply
#13
Wile E
Power User
qubit said:
You're very negative, aren't you? Sheesh.

It's not "his fault". :rolleyes: Ivy Bridge has been engineered to use significantly less power than Sandy Bridge: it fits into a 77W power envelope instead of 95W. Therefore, it should run cooler at stock not hotter, simple as that.

Having the CPU idle at 50C and then hit 75C+ on running a standard 3D Mark benchmark points to a fault somewhere. Perhaps the mobo is feeding it too much voltage? Could be a lot of things, but the fact remains that the CPU is running too hot and he's correct to flag it.
It's not too hot until it hits 105C. Under that it's just it being too hot for a user's preference.
Posted on Reply
#14
speedpc
I agree with you :respect:, I was just stateing that with a stock cooler at stock speed you would think it would run cooler. I've never had a cpu go over 60C under load on air. this Ivy is hitting that almost at idle.

Wile E said:
It's not too hot until it hits 105C. Under that it's just it being too hot for a user's preference.
Posted on Reply
#15
xenocide
qubit said:
It's not "his fault". :rolleyes: Ivy Bridge has been engineered to use significantly less power than Sandy Bridge: it fits into a 77W power envelope instead of 95W. Therefore, it should run cooler at stock not hotter, simple as that.

Having the CPU idle at 50C and then hit 75C+ on running a standard 3D Mark benchmark points to a fault somewhere. Perhaps the mobo is feeding it too much voltage? Could be a lot of things, but the fact remains that the CPU is running too hot and he's correct to flag it.
IVB was designed to use less power, but it's also on a smaller node, and as someone around here was pointing out for weeks before any IVB CPU's launched, it basically meant it would run hotter with the exact same elements as SB.

I am with Shibdib on this one, I call BS. There's no way it would have hit 75C+ without either a poor\incorrect configuration and\or an incorrectly configured BIOS. There's also just the off chance he got a bad CPU, it happens often enough. I just don't find it very likely a freshly installed CPU--even without aftermarket coolers--would hit almost 80c without some kind of problem on the user end.
Posted on Reply
#16
qubit
Overclocked quantum bit
xenocide said:
IVB was designed to use less power, but it's also on a smaller node, and as someone around here was pointing out for weeks before any IVB CPU's launched, it basically meant it would run hotter with the exact same elements as SB.

I am with Shibdib on this one, I call BS. There's no way it would have hit 75C+ without either a poor\incorrect configuration and\or an incorrectly configured BIOS. There's also just the off chance he got a bad CPU, it happens often enough. I just don't find it very likely a freshly installed CPU--even without aftermarket coolers--would hit almost 80c without some kind of problem on the user end.
I don't think you quite read my post properly. I also said that there had to be a fault somewhere and that these temperatures weren't normal. ShiBDiB simply said that there's nothing wrong with those temperatures, which is something else entirely.

Here's the pertinent bit of my post:

qubit said:
Having the CPU idle at 50C and then hit 75C+ on running a standard 3D Mark benchmark points to a fault somewhere. Perhaps the mobo is feeding it too much voltage? Could be a lot of things, but the fact remains that the CPU is running too hot and he's correct to flag it.
Posted on Reply
#17
speedpc
I received a new (different style to) Intel heatsink/fan today and the temps at idle went way down and under load the highest i've seen is 50C. Must have been the fan all along :)
Posted on Reply
#18
EarthDog
qubit said:
I don't think you quite read my post properly. I also said that there had to be a fault somewhere and that these temperatures weren't normal. ShiBDiB simply said that there's nothing wrong with those temperatures, which is something else entirely.

Here's the pertinent bit of my post:
Chances are something is wrong............

1. ...with the mount of his cooler.
2. ... its a VERY leaky chip which would make me want it. :p
Therefore, it should run cooler at stock not hotter, simple as that.
Actually... its not. You may not have thought of..........

1. The die is smaller than SB and doesnt have the same amount of space to dissapate the heat.
2. Have you seen die pics as in the guts/litho? Regardless, the iGPU is much larger on IB than SB so smaller die and smaller area to pack the cpu cores in compounds the heat issue.

If you put all that together with basic thermodynamic theory, it should lead you to believe that even though does use ~18W less, its also trying to dissapate heat through a MUCH smaller area therefore making it run hotter.
Posted on Reply
#19
cadaveca
My name is Dave
EarthDog said:
Have you seen die pics as in the guts/litho? Regardless, the iGPU is much larger on IB than SB so smaller die and smaller area to pack the cpu cores in compounds the heat issue.
Like this? :D


IVY:




SB:

Posted on Reply
#20
EarthDog
Spot on sir!!! Now slap a 2600k die/litho shot for comparison.

Looks like the iGPU here is 33% of the die, vs the 2600k which looks to be 25% of the die.
Posted on Reply
#21
cadaveca
My name is Dave
EarthDog said:
Spot on sir!!! Now slap a 2600k die/litho shot for comparison.

Looks like the iGPU here is 33% of the die, vs the 2600k which looks to be 25% of the die.
Personally, i think the temps sensors are FUBARRED..intentionally. Either way, it's not very important, iMHO, it is what it is, and there's no way to change it without voiding warranty. IF you find it an issue, IMHO, buy the Intel OC warranty for $30, clock the crap out of it, and rest easy. ;)
Posted on Reply
#22
EarthDog
I dont have a clue. Speculation everywhere and nothing to prove anything. I dont care either way if my CPU is running at 85C if it can handle it (which both IB and SB of course can) or 65C. its going to crunch the same 5 years from now either way.

Sad that I have IB and its in my review/benching rig and not using it elsewhere. :(
Posted on Reply
Add your own comment