Home / Component / APU / Intel Core i7-6700K ‘Skylake’ de-lidded: small die and NGPTIM found

Intel Core i7-6700K ‘Skylake’ de-lidded: small die and NGPTIM found

Although Intel Corp.’s upcoming Core i7-6700K “Skylake-S” chip will likely become overclockers’ new darling, it looks like it will have the same drawback as currently available mainstream microprocessors with unlocked multiplier: inefficient internal thermal interface.

As reported, Intel Corp. started to ship production ready full-speed engineering samples of its next-generation desktop microprocessors code-named “Skylake-S” to its partners in early June. Thanks to the fact that the new chips are now available relatively widely (but not commercially), more information about them is getting revealed these days. According to a recent discovery by a Coolaler forum member, the new Intel Core i7-6700K processors will allegedly continue to use thermal paste as their internal thermal interface between die and heat-spreader (if the blurry photo below is to be believed), a bad news for overclockers.

intel_skylake_s_6700k_delidded

In the last 15+ years virtually all desktop microprocessors from Intel featured metallic heat-spreaders (which many call lids) on their dies to protect the latter from damages and enable more efficient heat dissipation. For many years, Intel soldered dies to heat-spreaders using an exceptional Indium-based alloy, which enabled very efficient transfer of heat from silicon to metal. However, starting from the code-named “Ivy Bridge” central processing units Intel replaced the solder with a layer of thermal paste, which is considerably less efficient than the previously used alloy. For high-end desktop (HEDT) microprocessors, such as Intel Core i7 Extreme “Haswell-E”, the world’s largest chipmaker continues to use solder, thus providing exceptional thermal conductivity.

Many overclockers complained that in order to unlock maximum overclocking potential of the Core i7-3770K and the Core i7-4770K microprocessors they had to remove the heat-spreaders (de-lid the chips) and change thermal interface. To improve overclocking potential of its Core i7-4790K and Core i5-4690K “Devil’s Canyon” processors last year, Intel started to use its next-generation polymer thermal interface material, which is not bad, but is not as efficient as commercially available high-end thermal compounds.

As it appears from the blurry photo published at Coolaler forums, Intel will not solder heat-spreaders to its mainstream processors for enthusiasts, even if such chips are aimed at overclockers and are designed to run at high clock-rates. Apparently, if one wants to overclock Intel’s new chips to the max, he or she will need to switch Intel’s NGPTIM thermal interface first.

The blurry photo also reveals that the die size of Intel’s “Skylake-S” microprocessor is not large at all, which is means that it will be easy to overclock it; but its thermal density will be high, which means that it will not be easy to cool it down.

Intel Core i7-6700K central processing unit integrates four cores with Hyper-Threading and features 4.0GHz frequency, 4.20GHz maximum Turbo Boost clock-rate, 8MB last-level cache, a dual-channel DDR3L/DDR4 memory controller with 1600MHz or 2133MHz support, 95W TDP, Intel HD Graphics 6000-series integrated graphics core as well as LGA1151 packaging. The chip is expected to hit the market in August.

Intel did not comment on the news-story.

Discuss on our Facebook page, HERE.

KitGuru Says: As it appears, Intel simply does not want to enable maximum overclocking potential for its mainstream microprocessors. Enthusiasts can now only hope that Intel will use some kind of second-generation NGPTIM to maximize overclocking potential. Unfortunately, large corporations are usually pretty conservative with their choice of materials…

Become a Patron!

Check Also

AMD Krackan Point APU appears in the Geekbench database

AMD's next-gen Ryzen AI 300 series APUs are on the horizon, and a fresh Geekbench …

11 comments

  1. is Broadwell still under NDA ? i cant find any benchmarks, reviews.. how it plays againt 4790k.. i dont care about Skylake.. DDR4 not worth it until late 2016.

  2. -> I left m­­y desk wo­rk and since then I earn $85 every h. …How I did it? I am working via internet from my home! My old work didn’t exactly make me happy ,so I was forced to try something different… After Two years…I say it was the best decision I ever made! Here’s what I do <-

  3. Everyone is dropping or skipping over Broadwell. Why is DDR4 not worth it?

  4. Because it offers no real performance gains for most people, but it is more expensive.

  5. yep, thats why..

  6. I thought Broadwell and Skylake were both DDR4 parts, does it mean we won’t see broadwell desktop cpu’s and we’re going straight to Skylake? or is Skylake really a re-branded Broadwell?

  7. you don’t follow the cpu news? and you know the code names like broadwell and skylake? that’s not very common. no, broadwell doesn’t support ddr4, and no, skylake is not rebranded broadwell, broadwell is tick and skylake is tock.

  8. I had not seen enough press regarding broadwell desktop parts to really know what it was,have they even been released?? If they have then why are skylake performance numbers only compared to devil’s canyon sku’s, it would be easy to suggest given the conspicuous absence if broadwell and irs comparisons that although skylake is supposed to be tock it is in fact the CPU broadwell was meant to be, isn’t it ?

  9. It’s only more expensive because there is little demand, once skylake is out and the manufacturers can scale up production price will drop like a rock.

  10. To be frank, it wouldn’t be difficult for Intel to brand Broadwell parts under the Skylake nomenclature to get them sold off and quickly EOL them. Most people don’t pay attention to codenames, they rather are all about brand models, gives them an idea of what to buy next.

    I just find it odd that the 6700K uses the same IGP nomenclature as Haswell, it has an IGP branded as “HD5000” and happens to use 40-eu. Granted, do what works, I just expected a different series name for everything related to Skylake. Intel bothered to debut Haswell mainstream and Ivy Bridge-E under the same branding: Core i 4000 series. The flagship Sandy-E should have been called Core i7 29x0X, Ivy-E at 39x0X, Haswell-E with 49x0X– but that didn’t happen.

  11. Skylake = Another disappointment from Intel.

    http://semiaccurate.com/2015/08/05/intel-plays-press-skylake-stupidity/

    Probably 2-5% in procesor performance increase and 20-30% in integrated
    graphics increase.

    Taking in account that no gamer looks for integrated
    graphics as everybody have decent graphics card/s so it stays at best 5%
    increase.
    Intel is forcing everybody to buy their low performance integrated
    graphics card which to gamer is like a spare part. Like I would buy another
    graphics card and kept it in attic and use it only when my AMD Radeon or
    Nvidia GPU would blow up. Not even then as if my AMD or Nvidia would
    got broken I wouldn’t use Intel iGPU just went to the shop and bought
    another AMD or Nvidia GPU. So Intel is scamming me into buying product
    which I will never use!

    Even worse Skylake has still only 4 cores by default! We had 4
    cores default for last 8 processor generations and 5 technology shrinks:
    65nm, 45nm, 32nm, 22nm and now 14nm. This is an outrage. Nehalem in
    2008 had 4 cores 8threads configuration(45nm) I had Core 2
    Quad(Kentsfield 65nm) on LGA 775 which had…4 cores! in March…2007!

    My point is that with each process shrink there should be more cores
    added: 65nm-4cores, 45nm-6cores,32nm-8cores,22nm-12cores,14nm-16cores as
    standard keeping the same price per each processor generation. And all
    that without integrated graphics card which no gamer needs. Intel puts
    iGPU to manipulate marketing saying that they
    achieved 30% increase over last generation. (but it is only 3-5%
    processor increase!) iGPU actually only lower yields and
    adds to price and thermal generation effectively lowering OC
    capabilities. iGPU has only sense in low power laptop/tablet solutions.

    Why they do not add more cores keeping same price? Because it would make
    a competition to their server business.

    Intel is laughing at us and we are stupid to buy their processors. They are milking us. Skylake is another joke!

    I haven’t bought any new processor since my sandy 2500K
    running at 4.7Ghz i.e. since 2011. I hoped for Skylake to be 6-8 core by
    default with 4-6core price range but now I know it is another Intel
    emberassment.The problem is that there is no real alternative. So the
    only strategy for us is to wait. Either they will do some changes in
    their policy or they will bankrupt. I will wait for processor with at
    least 10 cores in price of todays 4 core to change. No more milking! Very
    disappointed.

    The
    point is we should have 16core processors on 14nm process BY DEFAULT
    already.(it goes for Intel, AMD and all other CPU makers)

    It is easy: if current Skylake 4 core is only 40% of CPU die space and 60%
    is integrated graphics then just remove iGPU and add 60% of space CPU
    cores(i.e. 6cores) = you end up with 10core CPU (maybe with a bit lower
    clock) within the same price range as 4 core Skylake Intel is feeding
    us. (also skylake die is very small it could easily be bigger with little price increase = 16 or 20 cores)

    Then when you need extra power and want to pay more you can
    calculate it for current Intel 6 core model = would be 16 core and
    current 8 core model would be 20core in the corresponding price ranges.

    And
    yes, I am aware that not all apps can use more cores(nondeterministic
    polynominal problems) but many other mathematical algorithms do. So the
    only way now to give us more processing power is add more cores! (hope
    new materials in the future will add more speed i.e. 1Thz but that is
    very far away from now) Developers would quickly utilize extra power.

    I hope everybody understand my point now. i.e. We could have 10 core Skylake in
    price of 4 core Skylake now and there is no technical problem to do
    so… only Intel’s policy. 🙁

    Skylake = Another disappointment from Intel.

    http://semiaccurate.com/2015/08/05/intel-plays-press-skylake-stupidity/

    Probably 2-5% in procesor performance increase and 20-30% in integrated
    graphics increase.

    Taking in account that no gamer looks for integrated
    graphics as everybody have decent graphics card/s so it stays at best 5%
    increase.
    Intel is forcing everybody to buy their low performance integrated
    graphics card which to gamer is like a spare part. Like I would buy another
    graphics card and kept it in attic and use it only when my AMD Radeon or
    Nvidia GPU would blow up. Not even then as if my AMD or Nvidia would
    got broken I wouldn’t use Intel iGPU just went to the shop and bought
    another AMD or Nvidia GPU. So Intel is scamming me into buying product
    which I will never use!

    Even worse Skylake has still only 4 cores by default! We had 4
    cores default for last 8 processor generations and 5 technology shrinks:
    65nm, 45nm, 32nm, 22nm and now 14nm. This is an outrage. Nehalem in
    2008 had 4 cores 8threads configuration(45nm) I had Core 2
    Quad(Kentsfield 65nm) on LGA 775 which had…4 cores! in March…2007!

    My point is that with each process shrink there should be more cores
    added: 65nm-4cores, 45nm-6cores,32nm-8cores,22nm-12cores,14nm-16cores as
    standard keeping the same price per each processor generation. And all
    that without integrated graphics card which no gamer needs. Intel puts
    iGPU to manipulate marketing saying that they
    achieved 30% increase over last generation. (but it is only 3-5%
    processor increase!) iGPU actually only lower yields and
    adds to price and thermal generation effectively lowering OC
    capabilities. iGPU has only sense in low power laptop/tablet solutions.

    Why they do not add more cores keeping same price? Because it would make
    a competition to their server business.

    Intel is laughing at us and we are stupid to buy their processors. They are milking us. Skylake is another joke!

    I haven’t bought any new processor since my sandy 2500K
    running at 4.7Ghz i.e. since 2011. I hoped for Skylake to be 6-8 core by
    default with 4-6core price range but now I know it is another Intel
    emberassment.The problem is that there is no real alternative. So the
    only strategy for us is to wait. Either they will do some changes in
    their policy or they will bankrupt. I will wait for processor with at
    least 10 cores in price of todays 4 core to change. No more milking! Very
    disappointed.

    The
    point is we should have 16core processors on 14nm process BY DEFAULT
    already.(it goes for Intel, AMD and all other CPU makers)

    It is easy: if current Skylake 4 core is only 40% of CPU die space and 60%
    is integrated graphics then just remove iGPU and add 60% of space CPU
    cores(i.e. 6cores) = you end up with 10core CPU (maybe with a bit lower
    clock) within the same price range as 4 core Skylake Intel is feeding
    us. (also skylake die is very small it could easily be bigger with little price increase = 16 or 20 cores)

    Then when you need extra power and want to pay more you can
    calculate it for current Intel 6 core model = would be 16 core and
    current 8 core model would be 20core in the corresponding price ranges.

    And
    yes, I am aware that not all apps can use more cores(nondeterministic
    polynominal problems) but many other mathematical algorithms do. So the
    only way now to give us more processing power is add more cores! (hope
    new materials in the future will add more speed i.e. 1Thz but that is
    very far away from now) Developers would quickly utilize extra power.

    I hope everybody understand my point now. i.e. We could have 10 core Skylake in
    price of 4 core Skylake now and there is no technical problem to do
    so… only Intel’s policy. 🙁