-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
Ivy Bridge Temperatures Could Be Linked To TIM Inside Integrated Heatspreader: Report
- Thread starter Twirlywoos
- Start date
I don't believe that jap thing. My ivy at 4.6ghz produces the following temps
Which are very close to this heatspreader-less cpu
Ignore its max temperatures, I was trying prime at 4.8 and 4.9ghz.
Your forgetting not all chips are equal, mine at 4.6 runs about 9 - 12c hotter than yours. Just because you got a good one doesnt mean all are well, some TIM on the chips might be better applied than others etc.
I don't believe that jap thing. My ivy at 4.6ghz produces the following temps
Which are very close to this heatspreader-less cpu
Ignore its max temperatures, I was trying prime at 4.8 and 4.9ghz.
To me it reads as they have replaced the TIM between the IHS and chip and it reduced temps by upto 20c, they do not say they removed the IHS.
Also your max temps of 80c are nowhere near close to 64c.
You say ignore the max temps, how can we when we see 80c, and if you were testing at different voltages and speeds then realtemp would have been reset surely?
Last edited:
I don't believe that jap thing. My ivy at 4.6ghz produces the following temps
Which are very close to this heatspreader-less cpu
Ignore its max temperatures, I was trying prime at 4.8 and 4.9ghz.
It doesn't say what they were using for stress testing, might have been Intel Burn Test which can be up to 10C hotter than prime, feel free to run it on yours and see what temps you get.
I think its easy isn't it ? the K series are OVERCLOCKING chips right ?
so take out Intel "warranty", overclock it with a decent cooler, stay within the specs whatever it is (1.35 or 1.4V ?) and if it keels over due to getting over 100C regularly - then your covered.
you're not using it outside of what its designed for
the K series are specifically for going above the multipliers that the non-K series are for - and you pay a premium for that previlege - so its the useage that its designed for
I'm going to ignore the temps for now
so take out Intel "warranty", overclock it with a decent cooler, stay within the specs whatever it is (1.35 or 1.4V ?) and if it keels over due to getting over 100C regularly - then your covered.
you're not using it outside of what its designed for
the K series are specifically for going above the multipliers that the non-K series are for - and you pay a premium for that previlege - so its the useage that its designed for
I'm going to ignore the temps for now
Well its quite simple, it was under full load as shown at 100% with 4.6Ghz.
I'll go do it again with IBT and repost when I get back from work.
But that could have been a ten second run to get to that for all we know
, I would be careful with IBT, lots of people say not to run it as it will hit 100c+ with it, just so you are aware.Yeah, please do it. That way I can sleep better at night knowing intel didn't do it on purpose hah.
Do it
If you try without the IHS make sure to get proper contact/pressure as someone supposedly tried that but with a standard air cooler that could barely touch the chip and therefore not effectively transfer heat.
Soldato
- Joined
- 25 Sep 2009
- Posts
- 10,256
- Location
- Billericay, UK
I think you right. That article looks and reads like a promo for Liquid Pro TIM but a pretty good one!
Is Liquid Pro rated amongst overclockers? In most threads people still plug TIM's like Arctic Silver 5.
Isn't the idea of the IHS to protect the die from being crushed by the heavy heatsinks?
The problem is simple to understand. Dissipating 77W of heat from a IB die of such small size is much harder than dissipating 95W of heat from a 210% larger SB die. The so called 3D and gapless transistors probably add to the problem.
The only way to properly reduce the heat will be to lower the TDP (make each transistor more efficient, or increase the diew size without increasing transistor count). It seems Intel have either been a little too greedy on shrinkage, or TDP is much higher than expected with IB.
The only way to properly reduce the heat will be to lower the TDP (make each transistor more efficient, or increase the diew size without increasing transistor count). It seems Intel have either been a little too greedy on shrinkage, or TDP is much higher than expected with IB.
AS5's still top notch as it spreads extra thin with a credit card, but you need to check that your cooler makes full contact. If it doesn't you're better off using something else with the blob method.
Liquid pro suffers the same issue, only works if both the CPU and cooler are flat otherwise you're better off using a blob of something else to fill the gap.
Liquid pro suffers the same issue, only works if both the CPU and cooler are flat otherwise you're better off using a blob of something else to fill the gap.
But that could have been a ten second run to get to that for all we know
I ran IBT at 4.7Ghz and it didn't go up in flames.
Why would I post a ten second run exactly? I am a little too old to be posting lies on the interwebs
Would you like a loaf of bread with my screenshot?!?!The temperatures were after a hour at that speed, I had it at 4.8ghz but was trying to find a nice compromise between power draw, heat and what not.
Will post more tomorrow when I'm back from work.
I ran IBT at 4.7Ghz and it didn't go up in flames.
Why would I post a ten second run exactly? I am a little too old to be posting lies on the interwebs
The temperatures were after a hour at that speed, I had it at 4.8ghz but was trying to find a nice compromise between power draw, heat and what not.
Will post more tomorrow when I'm back from work.
I all ways believe random guys on the internet
Loaf of bread not kneaded
as I have one of my own, thanks though.So you ran IBT, did you do it at maximum stress for 10-20 runs? would be interested to see that and of course the temps that go with it.
True. This is more of a step forward for mobile processors than desktop.