Tuesday, September 5th 2017

Intel Skylake-X HCC CPU Delidded by Der8auer, also not Soldered

Overclocking poster-boy Der8auer has seemingly gotten his hands on some early samples of Intel's Skylake-X high core count (HCC)HEDT CPUs. The upcoming 12 to 18-core enthusiast-class CPUs are being launched on the same X299 platform on socket LGA 2066 that Intel has already launched 4 (Kaby Lake-X), 6, 8 and 10-core parts already, and are supposed to bring Intel towards a level playing field - and then some - with competitor AMD's Threadripper CPUs, which boast of up to 16 cores.

From this delidding process with Der8auer's own delidding tool, Delid-Die-Mate-X, seems to result a die that is much larger - as expected - than Intel's 10-core i9-7900X. At the same time, it seems that Intel is still opting, again, for not soldering its enthusiast-targeted CPUs, which would result in better temperatures and, potentially, overclocking potential. The fact that Der8auer managed to delid the i9-7920X and didn't recommend against doing it likely means that there is minimal risk of damaging your CPU while subjecting it to this process. This is something the renowned overclocker did do when he recommended that users shouldn't delid their Ryzen or Threadripper CPUs looking for better temperatures, since the fact that these were soldered would likely result in both catastrophic damage and a much diminished chance of operating temperatures improvement through the application of special purpose thermal compounds. The Facebook post from Der8auer with the delidded 7920X likely serves as an appetizer for an upcoming delid video on YouTube, as has been the overclocker's MO.

Sources: Der8auer's Facebook, via Overclock 3D
Add your own comment

171 Comments on Intel Skylake-X HCC CPU Delidded by Der8auer, also not Soldered

#1
biffzinker
INSTG8R said:
What I have always wondered as it was Devil's Canyon they made claims of using better TIM well has every processor since been using that TIM or was it just a one time thing for DC and the rest use the bog standard stuff. I mean I hope at least they would with the HEDT if they have stopped soldering those as well.
Ah yes Intel Corp.’s NGPTIM next-generation polymer thermal interface material plus revamped power supply circuity was suppose to lead to improved overclocking.

One sample i5 4690K


Source for graphs: KitGuru
Posted on Reply
#2
INSTG8R
biffzinker said:
Ah yes Intel Corp.’s NGPTIM next-generation polymer thermal interface material plus revamped power supply circuity was suppose to lead to improved overclocking.

One sample i5 4690K


Yeah that is kinda what I figured. It was just a"minor effort" after the 4770K issues. I'm sure they've gone back to whatever they were buying in 45gallon drums for everything...I mean mine is idling at 27C and never passes 60C under load so it's not "terrible" but as I said I'm not pushing it either. It's just sad that it's really nothing cost wise to have not continued using superior solder at least on their HEDT chips...
Posted on Reply
#3
Vayra86
cadaveca said:
That's an affect (no, not a typo) of the mentality represented by users that do so, that new users that start into this "overclocking" hobby adopt. I mean, after all, that's what overclocking really used to be, before it was turned into a selling point. And because it's an affect (again, not a typo), it might not be the actual right approach. So when it was turned into a selling point, this affect was considered, and now we have things like every single multiplier, PER CORE of a CPU, already having a pre-programmed VID.

And while you may think Intel sells "K" SKU CPUs for overclocking, again, this is an affect of the community. "K" SKUs are ALREADY OC'ed for you. The power and cooling needs are increased already, and that should have been the first sign that THAT is what Intel is supporting... they have pre-OC'd the "K" SKUs for you. That's why many of these chips come without a cooler... you need to buy better than what Intel provides with the other chips, and the power increases considerably 65W - 95W is a near 50% increase in power consumed.

Unfortunately, marketing and reviewers have portrayed this differently than what it really is, but the truth of the matter is printed right on the box. Sure, you do also get "unlocked" multipliers, but do keep in mind that those multis already have VIDs programmed. If Intel truly wanted you to OC as you'd like to define it, they'd not have to waste their time doing things like programming the multipliers above stock with VIDs.
^This.

Its the same everywhere. Intel has been clock bumping the same junk for years now, and they stick a K on some of their chips to cater to a small percent of the market that wants to tinker with it - and they let them pay a premium for it which, combined with the required cooling solution, makes the net perf/dollar metric worse than any of their non-K options - the only gain you can extract is absolute performance. Nvidia: GPU Boost 3.0 eats almost all the headroom that's left in there at a given power target - its so horribly efficient at it, that it even has a positive effect on the perf/watt, and the AIBs then push the base clock a bit further, and we PAY for that OC being done for us.

Its so ironic really, TIM is the least of our problems when it comes to overclocking. Especially because those that really want the last 1% of performance are doing the delid anyway, solving the problem themselves. The rest of the world just wants someone else to do the work for them.

The whole thing is blown way out of proportion.
Posted on Reply
#4
Hood
biffzinker said:
So does that make @P4-630 a Intel fanboy then I presume since he named himself after the Pentium 4 630?

Probably.
Posted on Reply
#5
Th3pwn3r
EarthDog said:
Welcome to the 1%...

As we said, most dont care. Its not a big deal. I like to push things to the limit and this frankly doesnt bother me. I mean, what is 200 mhz more IF IM LUCKY (we went over it at this end of overclocking you arent gaining much..even gave my results).

Again, even with the 'not optimal tim' its still overclocking way further than the competition.

I also agree it has very lite to do with the upgrade cycle by gaining 200 mhz with better TIM. Seriously, 200 mhz isnt going to let you keep your processor for much longer...
Only real difference could be between a processor being a bottleneck or not, depends on which one we're talking about. 200mhz is a huge deal for a Pentium III
Posted on Reply
#6
EarthDog
200 Mhz doesn't make a PIII playable today. Maybe it did then, but, well, this is now. :)
Posted on Reply
#7
Th3pwn3r
EarthDog said:
200 Mhz doesn't make a PIII playable today. Maybe it did then, but, well, this is now. :)
That's the point, 200mhz won't make or break anything. It's more for bragging rights.
Posted on Reply
#8
Valantar
Th3pwn3r said:
80c doesn't leave much headroom though. If your processor is running at 80c stock clocks you can pretty much forget about any overclocking.
Considering those temps were claimed for a 10-core system running at 4.5GHz on all cores, I doubt there's much room for overclocking left at all, no matter what creates the bottleneck. As such, 80C is perfectly acceptable - I'd say it's pretty damn good. And it sure isn't going to damage anything (although voltage might, of course, but that's another story entirely).
Posted on Reply
#9
trparky
cadaveca said:
And while you may think Intel sells "K" SKU CPUs for overclocking, again, this is an affect of the community. "K" SKUs are ALREADY OC'ed for you. The power and cooling needs are increased already, and that should have been the first sign that THAT is what Intel is supporting... they have pre-OC'd the "K" SKUs for you.
Yep. You basically said what I said earlier in this thread. The K chips are basically already pushed to the limits at the factory aka pre-overclocked.
Posted on Reply
#10
EarthDog
trparky said:
Yep. You basically said what I said earlier in this thread. The K chips are basically already pushed to the limits at the factory aka pre-overclocked.
But they arent... we said that. :)

They easily reach past their boost and many past 5ghz (7700k/7740k) with good cooling. AMD doesnt get past their own boost. If intel is at the limit, you must feel amd is past its limit?

7900x will overclock 1 ghz and then some...

Th3pwn3r said:
That's the point, 200mhz won't make or break anything. It's more for bragging rights.
Its my point as well...not sure why you brought the p3 up if you agreed. Seemed contrary.

Edit: i get you now... yeah. My reply to that is simply anyone can find a rare use case to disprove anything. :)
Posted on Reply
#11
trparky
I guess that we have a different idea what limits are. The K chips are pushed to what I would call "sane" limits. If you take a K chip and further overclock it you generally need far better cooling than your average person would ever think of putting into their rigs.
Posted on Reply
#12
EarthDog
I guess we do :). Now that it is clearly defined, I still wonder what you think AMD's chips are. ;)

Seems we also have a different definition of what 'far better cooling than your average person would ever think of' is too. Many 7700K/7740K will reach 5GHz+ with an AIO. Many average people use AIOs these days. :)

Hell, the 4.5GHz on my 7900X, a 1.2GHz overclock from base, I can cool with an AIO............ on that 'less than optimal' TIM. ;)
Posted on Reply
#13
trparky
AMD's chips are pushed to the point where... They're balls to the walls.

I thought that at 5 GHz you needed custom cooling and no AIO would be able handle it.
Posted on Reply
#14
Th3pwn3r
Valantar said:
Considering those temps were claimed for a 10-core system running at 4.5GHz on all cores, I doubt there's much room for overclocking left at all, no matter what creates the bottleneck. As such, 80C is perfectly acceptable - I'd say it's pretty damn good. And it sure isn't going to damage anything (although voltage might, of course, but that's another story entirely).
Yep, cooler is better but I'm not sure who said 80c would damage something. It's not a great starting point for over clocking though. There are plenty of people near or above 80c when shooting for 5ghz. After a delid and new TIM and sealant temps drop and higher clocks are obtained on a lot of 7700k processors out there.

EarthDog said:
But they arent... we said that. :)

They easily reach past their boost and many past 5ghz (7700k/7740k) with good cooling. AMD doesnt get past their own boost. If intel is at the limit, you must feel amd is past its limit?

7900x will overclock 1 ghz and then some...


Its my point as well...not sure why you brought the p3 up if you agreed. Seemed contrary.

Edit: i get you now... yeah. My reply to that is simply anyone can find a rare use case to disprove anything. :)
Yep, to each his or her own. I think too many people on here at grasping at straws just to be angry. I'm always on the fence for anything.
trparky said:
AMD's chips are pushed to the point where... They're balls to the walls.

I thought that at 5 GHz you needed custom cooling and no AIO would be able handle it.
Someone posted a build on Reddit saying he has 5GHZ stable on his 7700k with just a 120MM Corsair AIO.
Posted on Reply
#15
Valantar
Th3pwn3r said:
Yep, cooler is better but I'm not sure who said 80c would damage something. It's not a great starting point for over clocking though. There are plenty of people near or above 80c when shooting for 5ghz. After a delid and new TIM and sealant temps drop and higher clocks are obtained on a lot of 7700k processors out there.
Again: we're not talking about 80c as a starting point for OC'ing, but as the end result. Those are two radically different things. Now, I'm not talking about 7700Ks, nor am I saying that 7900Xs don't run hot, I'm simply pointing out the complete lack of logic of saying that a heavily overclocked 10-core running at 80c with all cores at 4.5GHz is a perfectly fine result.
Posted on Reply
#16
trparky
Valantar said:
eavily overclocked 10-core running at 80c with all cores at 4.5GHz is a perfectly fine result.
You see, they're not supposed to be running at those speeds; they were never designed to me running at those speeds. Yes, Intel gives users the ability to overclock their enthusiast chips but the conditions that occur after user overclock them is completely up to the users to handle. Intel knows the limits of their chips, the fact that users overclock them is completely up to them.

Look at it this way, back in the days of Ivy Bridge their chips were clocked in the mid-3 GHz range and you were able to get some amazing overclocks out of them. I myself were able to get an extra 1 GHz out of it. Why? Because at the time the chips were clocked lower than what is needed today. Fast forward to today and Intel knows what is needed and already gives it to you in the form of the K chips, pre-overclocked chips. Any more overclocking than Intel has already done for you is completely up to you. Think of the K chips like the pre-overclocked video cards you often see. Yep, already overclocked, anymore overclocking and you risk damage to it.
Posted on Reply
#17
Th3pwn3r
Valantar said:
Again: we're not talking about 80c as a starting point for OC'ing, but as the end result. Those are two radically different things. Now, I'm not talking about 7700Ks, nor am I saying that 7900Xs don't run hot, I'm simply pointing out the complete lack of logic of saying that a heavily overclocked 10-core running at 80c with all cores at 4.5GHz is a perfectly fine result.
Explain to me like I'm 5 please(ELI5).
Posted on Reply
#18
StrayKAT
trparky said:
You see, they're not supposed to be running at those speeds; they were never designed to me running at those speeds. Yes, Intel gives users the ability to overclock their enthusiast chips but the conditions that occur after user overclock them is completely up to the users to handle. Intel knows the limits of their chips, the fact that users overclock them is completely up to them.

Look at it this way, back in the days of Ivy Bridge their chips were clocked in the mid-3 GHz range and you were able to get some amazing overclocks out of them. I myself were able to get an extra 1 GHz out of it. Why? Because at the time the chips were clocked lower than what is needed today. Fast forward to today and Intel knows what is needed and already gives it to you in the form of the K chips, pre-overclocked chips. Any more overclocking than Intel has already done for you is completely up to you. Think of the K chips like the pre-overclocked video cards you often see. Yep, already overclocked, anymore overclocking and you risk damage to it.
I'm just sticking stock and/or turbo myself. This whole de-lidding thing convinces me even more. I don't run anything that needs more than 4.5 anyways. Secondly, I don't want a big double tower to cool it. The lengths I have to go to OC are absurd. Should have never bought a K series probably.
Posted on Reply
#19
cadaveca
My name is Dave
Valantar said:
I'm simply pointing out the complete lack of logic of saying that a heavily overclocked 10-core running at 80c with all cores at 4.5GHz is a perfectly fine result.
And why isn't it perfectly fine? The CPU can go higher, both in speed and temperatures. It could go lower in temperatures, if I wasn't using a fairly cheap AIO cooler. While many want 5 GHz CPUs, for me, the optimal speed for Intel's CPU design since Sandybridge has been 4.6 GHz, and until they change their core design, that won't change.
Posted on Reply
#20
Th3pwn3r
StrayKAT said:
I'm just sticking stock and/or turbo myself. This whole de-lidding thing convinces me even more. I don't run anything that needs more than 4.5 anyways. Secondly, I don't want a big double tower to cool it. The lengths I have to go to OC are absurd. Should have never bought a K series probably.
It's not that my much work or hassle.

Check out the job I did yesterday-
Posted on Reply
#21
Valantar
cadaveca said:
And why isn't it perfectly fine? The CPU can go higher, both in speed and temperatures. It could go lower in temperatures, if I wasn't using a fairly cheap AIO cooler. While many want 5 GHz CPUs, for me, the optimal speed for Intel's CPU design since Sandybridge has been 4.6 GHz, and until they change their core design, that won't change.
I apparently managed to turn an "isn't" into an "is" while writing that. So much for making myself clear, I guess. It is absolutely a perfectly fine temperature, which was my whole point (and what others were arguing quite strongly against).
Posted on Reply
Add your own comment