der8auer and Caseking ready up tweaked Core i7-8700K CPUs

Seemingly everyone wants to go for a swim in the deepest reaches of Coffee Lake with Intel's range-topping Core i7-8700K. The chip has been pretty hard to find so far, but der8auer and his employers at German retailer Caseking have some very special i7-8700K's in stock for those willing to pay premium prices. Caseking's chips are de-lidded, binned, and at the most extreme level, topped off with an aftermarket heat spreader made of 99.9% silver.

Image source: der8aur's Facebook page

Caseking is offering binned chips guaranteed to hit 4.8 GHz to 5.2 GHz in three different versions, all etched with der8auer's and Caseking's logos. Chips labeled Advanced have been de-lidded and have a stock IHS, Pro versions are de-lidded with a polished stock IHS, and Ultra edition chips have had the stock IHS knocked off and replaced with an aftermarket all-silver unit. Each CPU is sold with notes about the voltage needed to reach the specified overclocked frequency. All chips meet their target speeds at no more than 1.4 V.

The binned and de-lidded chips are listed on Caseking's website right now. The only version in stock at the time of writing is the 5.1 GHz bin of the Advanced series chips for 580€ (about $544 without VAT). The least-expensive option is the 4.8 GHz Advanced chip at 440€ (approximately $413). The top of the range is the 5.2 GHz Ultra that goes for a heady 870€ (a little more than $816). For reference, Caseking offers tray versions of the stock Core i7-8700K processor for 325€ (about $305).

Comments closed
    • Mr Bill
    • 2 years ago

    A [url=https://www.1-act.com/vapor-chambers-2/<]vapor chamber heat spreader[/url<] would move heat more rapidly than silver and can be bonded directly to electronics. Not sure if that implies bonding directly to dies.

    • denstieg
    • 2 years ago

    Isn’t it wierd that a “stock” chip is unobtainable but caseking got enough to bin them and have more stock than anyone else, somthing wrong in the German import channels?.

    • Air
    • 2 years ago

    I’ll wait for the diamond version. Silver is only good enough for casual overclockers.

    • Chrispy_
    • 2 years ago

    I would have thought that making the IHS thinner was more beneficial to making it out of soft silver.

    Thermal conductivity isn’t really a factor that matters when the baseplate of your cooler is only copper, not silver. What [b<]*does*[/b<] matter is the sheer amount of material between the cold water in the CPU waterblock and the silicon die. A thinner IHS with a thinner layer of paste/solder/goo and a thinner waterblock. All of the material between the (relatively) cold water and the silicon die acts as a heat buffer. It's good for absorbing rapid temperature fluctuations in the die, but also affects the speed at which the wet side of the CPU block increases in temperature, which means there's a delay in getting that extra heat energy away from the CPU and into the water.

      • BurntMyBacon
      • 2 years ago

      [quote=”Chrispy_”<]I would have thought that making the IHS thinner was more beneficial to making it out of soft silver.[/quote<] I believe this is correct in most cases (will discuss below). Specifically, in these designs, I don't think the silver IHS will help much, but won't make things worse either assuming the silver IHS is the same thickness as the copper version it replaces. Using a significantly thinner IHS does have the caveat that you need to modify the mounting bracket from most coolers to use it. If you are going to go to that much trouble, you might want to consider running without an IHS (Like a Boss). [quote="Chrispy_"<]Thermal conductivity isn't really a factor that matters when the baseplate of your cooler is only copper, not silver.[/quote<] If I understand my theory correctly, you are half correct. Silver (vs copper) does not increase the effective thermal conductivity of the IHS to heatsink interface (assuming copper heatsink) given that the heatsink covers the entire IHS in both cases. However, given that the IHS is much larger than the silicon die itself, there are always temperature gradients across the IHS that make the effective transfer zone to the heatsink smaller than the full size of the IHS (I.E. not as much heat is transferred at the corners of the IHS as in the middle). The fact that silver has higher thermal conductivity than copper allows the heat from the die to spread more quickly in the horizontal direction on the IHS creating a larger effective surface area for transfer. Keep in mind, though, that silver is only something like 7% more thermally conductive than pure copper. [quote="Chrispy_"<]All of the material between the (relatively) cold water and the silicon die acts as a heat buffer. It's good for absorbing rapid temperature fluctuations in the die, but also affects the speed at which the wet side of the CPU block increases in temperature, which means there's a delay in getting that extra heat energy away from the CPU and into the water.[/quote<] I'm not sure I'm picking up what you're putting down here. While extra material would slow the speed at which the initial change in heat energy finds its way to the water, assuming the source is significantly hotter than the water (and no insulators are present) the heat energy will flow away from the CPU and towards the water, stabilizing at about the same temperature as it otherwise would have. Heat will not flow back into the CPU unless the CPU somehow gets cooler than the IHS. While the heat does take longer to get to the water, it is not sitting stagnant next to the CPU either. It is still consistently moving to cooler materials (away from the CPU). The copper to air or water interface is far more thermally resistive than a few mm of copper or silver. After the initial delay, the heat should be delivered to the water as fast as the water can remove it (given the waterblock design). The biggest issue in transferring heat is interfaces. I've worked with many conductively cooled embedded systems that are completely sealed and cooled at the surface. You'd be surprised how well it works with distances several orders of magnitude greater than the difference in IHS thickness. Every time I saw an unexpected temperature delta between the card and the cooler, it could be traced to a bad interface. Obviously, air is a poor thermal conductor, so even if we have to use thermal pastes that have two orders of magnitude less thermal conductivity than our heatsink, we need to get the air out of the interface. Even the indium solder has only about 1/5 the thermal conductivity of copper and a little more than 1/2 the thermal conductivity of the silicon die. It is these layers that are considered insulators (relative to the rest of the thermal conduction path) and these need to be as thin as possible to limit how much they impede heat flow. Making mating surfaces as flat as possible and using materials that can maintain that flatness (silver, being soft, may warp over time in extreme conditions), using no more thermal interface material than necessary, using thermal interface material that doesn't break down and loose thermal conductivity over time, and using good mounting techniques (often with higher mounting pressures) to maintain better contact are some ways to improve the interface thermal conductivity. Unfortunately, Intel has been using thinner substrate, poor mounting (read thick layers of thermal interface material), and thinner PCBs which make this interface less than ideal. While poor mounting of the IHS can potentially be fixed by deliding, the thinner PCBs make it difficult to achieve the thinnest interface possible. Indium solder would be a good way to mate these surfaces if only the silicon die weren't prohibitively thin (read: Will break under normal expansion scenarios). About the best thing you can do is polish the base of your heatsink as flat as you can and mate it to the silicon die directly to remove an unnecessary interface. You just have to be extremely diligent to avoid putting too much mounting pressure on the thinner PCB. A proper shim should be used to apply pressure to the outer parts of the PCB. It both evens out the pressure (big deal for Skylake) and makes necessary pin contacts that won't occur with pressure on the die alone. [i<]Edit: Spelling and Grammar[/i<]

        • Chrispy_
        • 2 years ago

        Your last paragraph kind of nails it:

        PCB too thin to save $0.01 per wafer,
        IHS made stronger to compensate,
        Lots of paste required since the flexible PCB creates variable gap.

      • not_a_gerbil
      • 2 years ago

      I’m annoyed that the die protector is marketed as a heat spreader by Intel. As you said, a thick layer, even if copper, is insulating the heat from a copper cooler.

        • Chrispy_
        • 2 years ago

        Yeah, IHS is definitely a misnomer and it just doubles the number of interfaces (as BurntMyBacon pointed out, interfaces are the weakest link, and Intel’s thermal paste is some of the worst).

        I used to use shims with my AMD chips. Nvidia and AMD also use exposed dies with shims in their GPUs. What’s wrong with using shims for CPUs again?

      • psuedonymous
      • 2 years ago

      [quote<]I would have thought that making the IHS thinner was more beneficial to making it out of soft silver.[/quote<] Too [i<]thick[/i<] a heatspreader and the thermal path would be dominated by the bulk conductance of the IHS materail (copper or silver). But I suspect that any current IHS is well, well below the optimum thickness in pursuit of reducing overall package thickness. The IHS is, after all, a heat[i<]spreader[/i<]. It's goal is to take a tiny die of ~100-150mm^2 contact area and mate it to a heatsink base that is ~1000mm^2. For example, if you used a direct-contact heatsink without an IHS (placed onto bare die) the heatsink would work [i<]awfully[/i<] having only 1/2 pipes in actual contact with the die. [quote<]All of the material between the (relatively) cold water and the silicon die acts as a heat buffer. It's good for absorbing rapid temperature fluctuations in the die, but also affects the speed at which the wet side of the CPU block increases in temperature, which means there's a delay in getting that extra heat energy away from the CPU and into the water.[/quote<] As long as that energy has been removed from the die (and by heating the IHS it has done by definition) then it is working completely effectively. It makes not a jot of difference [i<]how fast[/i<] a particular joule of emitted energy makes its way from the die where it is generated to the water (and eventually air), just that it exits the die itself in a timely manner. Conduction time does not matter in the slightest.

    • techguy
    • 2 years ago

    I appreciate Der8auer’s contribution to the enthusiast PC community, and I have one of his Delid Die Mate X’s on order, but this product line has no appeal to me. People can spend their money however they like but you can get a de-lidded and binned chip for way less than this by:
    a) doing it yourself (not recommended with Skylake-X)
    b) ordering from siliconlottery (even if you live in Europe)

    • Gadoran
    • 2 years ago

    8700K performs enough at stock speeds with acceptable temperatures, i don’t see reasons to purchase these top binned SKUs. Anyway…..

    • joselillo_25
    • 2 years ago

    This looks like aftermarket car business now, like brabus and all this companies , who said that car analogies were dead?

    • AMDisDEC
    • 2 years ago

    LOL, what a tremendous waste of time and money.
    There are definitely far too many over privileged Martin Shkrel brats in this world.

    • Krogoth
    • 2 years ago

    This is pure e-ricing.

    Die-hard overclockers just get rid of the headspreader and run the chip naked.

      • Firestarter
      • 2 years ago

      bring back the shims!

      • DrDominodog51
      • 2 years ago

      No. LN2 pots rarely support mounting directly to the die, and those that do are (relatively) expensive and hard to find.

      Extreme overclockers replace the the stock paste under the heatspreader with [url=http://thermal-grizzly.com/en/products/16-kryonaut-en<]Kyronaut[/url<], and put the heatspreader back on.

        • RAGEPRO
        • 2 years ago

        You probably meant [url=http://www.thermal-grizzly.com/en/products/26-conductonaut-en<]Conductonaut[/url<], chief. Kryonaut is a fine paste, but it's poor compared to Coollaboratory's stuff, or Conductonaut.

          • Nestea_Zen
          • 2 years ago

          kryo
          Temperature -250 °C / +350 °C

          Conductonaut
          Temperature 10 °C / +140 °C

          even the name implies it’s meant for sub zero temperatures.
          „Kryo” – the greek word for “cold” – also found in the german word for “cryoengineering”, suggests that this thermal grease was created especially for cryogenic applications – for the true “Kryonauts” amongst the “extreme overclockers”.

    • DPete27
    • 2 years ago

    $500 for 10% OC….. no thanks.

      • AMDisDEC
      • 2 years ago

      As Barnum said, there’s a fool born every minute, but updated today it’s, every second.
      And YES, he DID say it. So don’t bother quoting WiKi’s misinformation.

    • DPete27
    • 2 years ago

    How do you pronounce this joker’s name?

      • Voldenuit
      • 2 years ago

      Since they’re German… ‘der – acht – auer’?

      • TravelMug
      • 2 years ago

      8 = B, so der Bauer.

        • Redocbew
        • 2 years ago

        If that’s correct, then he called himself “the farmer”? There’s got to be a story behind that.

          • TravelMug
          • 2 years ago

          Yeah, I never looked up any story, but that would be one of the meanings. Not the same in this case as “the farmer” as in English, “Bauer” is more of a derogatory term to describe someone as a less sophisticated person, a bit like “culchie” in Irish-English. Can also of course mean “the builder” from “bauen” (to build) judging from his work, but that’s a bit weird sounding unless he was going for that in the first place. “the farmer” sounds more self-deprecating though 🙂

    • Wirko
    • 2 years ago

    At these prices I’d expect the heatsink to be soldered.

      • Beahmont
      • 2 years ago

      You can’t solder chips this small on substrates this thin. Over time as the ship heated and cooled it would buckle and crack.

        • Waco
        • 2 years ago

        Source?

          • smilingcrow
          • 2 years ago

          HP!

          • Beahmont
          • 2 years ago

          [url<]https://overclocking.guide/the-truth-about-cpu-soldering/[/url<] It's not just chip size, it's also that Skylake and beyond went to a much thinner substrate.

            • Waco
            • 2 years ago

            Which was to save a few bucks per CPU.

            Older smaller chips were soldered without issue. I’m not buying off on this just yet.

            • NTMBK
            • 2 years ago

            No, they thinned the substrate to reduce Z-height on mobile chips. Fractions of a millimeter mean thinner Macbooks.

            • Waco
            • 2 years ago

            …we aren’t talking about mobile chips (which typically don’t have an IHS). The substrate is entirely different on the desktop sockets.

            • psuedonymous
            • 2 years ago

            Desktop and notebook chips get laid down onto the same substrates. If you start to mix-and-match that sort of thing you then need to start maintaining different signalling hardware for the ULP and higher-TDP lines (becuase signalling is now taking a different oath from die to main board). This means you cannot bin the same die across power envelopes, but need to fab dedicated dies for each target platform and bin within that platform.

            • Waco
            • 2 years ago

            The desktop and mobile chips are on different substrates, though. That’s why the mobile chips aren’t socketed and are generally BGA versus LGA.

            • BurntMyBacon
            • 2 years ago

            I think you are confusing the silicon substrate for the PCB. They certainly use different PCBs, but even if the chips are cut from different wafers, the fact that both mobile and desktop chips are manufactured in the same facility and on the same process suggests that the silicon substrate thickness is going to be the same.

            • Waco
            • 2 years ago

            Bah. Indeed I was. Nevermind. 🙂

        • iBend
        • 2 years ago

        And why AMD can solder their IHS, but intel can’t?

          • Beahmont
          • 2 years ago

          Because their chips are larger and have a thicker substrate? Duh.

            • iBend
            • 2 years ago

            And Intel can make their chip larger and use thicker substrate if they want to, but why they didn’t?

            • Waco
            • 2 years ago

            …because that would cost more money. 😛

            • BurntMyBacon
            • 2 years ago

            I can’t say for certain that using a thinner substrate doesn’t have inherent benefits to the process, but in the past, most of the electrical activity was so close to the surface of the substrate that I’m having a hard time believing this is the case. The more likely driver is that using a thinner substrate nets Intel more wafers per ingot, reducing internal costs. Lack of indium solder also nets them $2 – $5 per processor as I hear it.

        • MOSFET
        • 2 years ago

        Like that bourbon aging in barrels at sea? (and now it’s magically worth 4x more, or $135/750mL)

    • DeadOfKnight
    • 2 years ago

    Uh…go and look at the rock bottom price of silver right now to see how much of a ripoff this is. For like the past 5 years the price of an ounce of silver has been under $20.

      • psuedonymous
      • 2 years ago

      You’re paying for the precision machining (because you need the dimensions right to avoid crushing the die or leaving too large a gap), not the raw feedstock. Make the same custom IHS out of Copper or Aluminium and the retail price would end up pretty much the same.

      It’s the same reason why short-run custom cases cost many times more than mass-produced stamped sheet cases, even if made of the same raw material. Machining costs, and the smaller your production run the more it costs per-unit.

    • Voldenuit
    • 2 years ago

    Sales motto: “Every core has a silver lining”.

      • jodiuh
      • 2 years ago

      Hee hee

    • K-L-Waster
    • 2 years ago

    Is silver actually better at heat dissipation, or is this just for bragging rights?

      • chuckula
      • 2 years ago

      It’s clearly targeting the mining crowd.

      The silver mining crowd.

      • Redocbew
      • 2 years ago

      It is better, but only slightly. The difference in conductivity between silver and copper is much smaller than the difference between copper and aluminum. I wouldn’t expect it to make much difference in practice.

        • AMDisDEC
        • 2 years ago

        Someone wasting their time playing games all day and wasting money to OC a CPU just to gain a few FPS wouldn’t know this but there is a drastic difference between electrical and thermal conductivity between metals.
        The thermal conductivity of pure silver is about 25% better than copper at ambient. Not to mention it has a higher valence as well which means it will retain these thermal/electrical properties longer.

          • Redocbew
          • 2 years ago

          I assume you’re talking about the valance of an element in relation to forming bonds and compounds. Thermal conduction is not a bond. It’s a transfer of energy, and if there’s any kind of electrical current passing between the IHS and heatsink, then you’ve got bigger things to worry about than thermals.

          I’m almost afraid to ask you to cite your source for the difference of 25%. I suppose if the copper had a lot of impurities you might be able to stretch the difference in conductivity between the two, but on even footing the difference isn’t nearly that much.

            • AMDisDEC
            • 2 years ago

            Of course there is a bond, between the two conducting materials and the transfer of heat is actually the transfer of energy from one plate to the other. The heat transfer is conducted by the excitation of free electrons in the valance shell.
            The higher the number of free electrons in the outer shell of the atom, the better it conducts electrical and thermal energies.

            As I stated, the +25% is at ambient T, or 20C.

            • Redocbew
            • 2 years ago

            Cue weirdest tangent ever #2.

            Yeah… not so much.

            • SomeOtherGeek
            • 2 years ago

            LOL. You’re a weird tangent magnet!

            • AMDisDEC
            • 2 years ago

            LOL, what a dunce!
            To say heat transfer is possible without a bond between surfaces and electrons aren’t involved in heat conduction cooling.
            You qualify for a Trump cabinet appointment.

            • Redocbew
            • 2 years ago

            [url<]https://www.youtube.com/watch?v=8J_F-GCJH6w[/url<]

            • AMDisDEC
            • 2 years ago

            LOL. Basic thermal dynamics 101 and the theory applies to human biology as well which explains why you display a learning deficient.
            [url<]https://s19.postimg.org/hq3lavt4j/Quantum.png[/url<]

            • brucethemoose
            • 2 years ago

            I’m probably screaming into the void, but… You seem to be confusing bulk conductivity and interface conductivity, as well as drastically oversimplifying both and pulling that 25% out of nowhere.

            • AMDisDEC
            • 2 years ago

            “Void” implies nothingness, Alb indicates white, or colorless.
            Physics is physics and you either know it or don’t.
            As example, there is good reason why melanin contains Copper and Superconducting Selenium for energy transfer.
            I can’t help being a know-it-all, as others for being know-nothings.

            • Waco
            • 2 years ago

            Please stop.

            • AMDisDEC
            • 2 years ago

            Less whining, and more fish and books.

            • chuckula
            • 2 years ago

            Pffttt… your sad pseudo-scientific babble clearly fails to account for the [url=https://en.wikipedia.org/wiki/Phlogiston_theory<]phlogiston[/url<] tensor field! Back to the rookery with you!

            • AMDisDEC
            • 2 years ago

            Sounds like the grave-robber calling thief.

            I suggest you look more carefully at 1800-1900s European science, including the mummy medicine market.

      • Duct Tape Dude
      • 2 years ago

      According to this it’s about 7% more thermally conductive than copper: [url<]https://en.wikipedia.org/wiki/List_of_thermal_conductivities[/url<] If only they made a diamond heatspreader...

        • AMDisDEC
        • 2 years ago

        They do, but not for gamers.

        • BurntMyBacon
        • 2 years ago

        Alternately, they should start fabricating the processors on diamond instead of silicon. It would go a long way to solving the heat density problem as the heat generating elements get smaller.

          • AMDisDEC
          • 2 years ago

          LOL, I just heard that the Trump administration has moved coal to the top EPA supported fuel source and they have renamed it to, magic rocks.
          Coal, oops I meant, magic rocks, are the precursor to diamond, right? So, according to the EPA, diamonds are Magical Magic Rocks, or MMRs.

      • Zizy
      • 2 years ago

      Yes and yes.

        • K-L-Waster
        • 2 years ago

        Fair point – not mutually exclusive.

    • the
    • 2 years ago

    A 99.9% silver heat spreader? That’s rather impressive and also expensive. Just how much does that weight exactly?

      • Waco
      • 2 years ago

      I can’t imagine it’s much better than the copper stock IHS…but style points matter to some I guess.

        • chuckula
        • 2 years ago

        I’m assuming [hoping] that somebody did the chemistry to make sure that whatever liquid-metal TIM they are using is copacetic with silver.

          • Waco
          • 2 years ago

          Ha, that’d be funny if not.

          WHY IS MY CPU MELTING????

            • psuedonymous
            • 2 years ago

            People already put eutectic TIMs directly onto the Die, which falls squarely into Bad Idea territory.

            • RAGEPRO
            • 2 years ago

            What makes you say that?

        • Takeshi7
        • 2 years ago

        Silver is actually the best thermally and electrically conductive element. Even better than gold.

        The reason electrical connectors are gold plated instead of silver plated is because silver oxidizes.

        [url<]http://www.engineeringtoolbox.com/thermal-conductivity-metals-d_858.html[/url<]

          • Waco
          • 2 years ago

          I’m well aware. I would be surprised if there’s a real difference for swapping out the IHS, though.

            • Takeshi7
            • 2 years ago

            Well according to that table silver should be up to ~15% better at conducting heat. I’d imagine these heatspreaders would be measurably better, but probably not worth their asking price.

            • Waco
            • 2 years ago

            Sure, but the limiting factor is rarely the conductivity of the IHS.

            Similar to thermal compound – there are some that are many times better at conducting than the regular grey goo…and they yield comparatively little in testing. An IHS is a lot thicker, but still, unless your heat sink is also solid silver with a kickass loop behind it, I just don’t see the point. 🙂

      • Mr Bill
      • 2 years ago

      $28 for one [url=https://www.amazon.com/2015-American-Silver-Dollar-Uncirculated/dp/B00RPI0X2U<]ounce[/url<], uncirculated!

        • DeadOfKnight
        • 2 years ago

        [url<]http://www.kitco.com/charts/livesilver.html[/url<]

          • Mr Bill
          • 2 years ago

          yes but using a silver dollar has so much more cachet.

          • SomeOtherGeek
          • 2 years ago

          You don’t own it unless it is in your hands…

Pin It on Pinterest

Share This