• 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.

I9 9900k

Distracted said:
Seems that way. The processors that get binned are separate stock according to Gibbo. So after binning even those that don't reach high clocks remain with the system builder people and don't get put back into the warehouse stock for selling on.

That is as much info as we have had about it.

So when you buy an oem from Ocuk it should be new, plain box, 1yr warranty, but with the intel starter software pack including killing floor 2, some adobe stuff and COD 4.
Click to expand...

Thx for the answer, much appreciated
 
Legend said:
I don't know what's going on... has buying a CPU now become a lottery as to whether it runs crazy hot or not?? Or is it simply early batch issues? Unacceptable whatever is happening.
Click to expand...

My understanding is a little different, the der8auer video is a good explanation of it and to be honest that's really the focus of his entire 9900k delidding video.

Basically there's 3 materials and each has a different level of conductivity. The more conductive the quicker heat will get through to whatever cooling you're using. There's the chip itself, there's the TIM and then there's the IHS. The IHS he explains is copper so it's high conductivity so that's not really the issue. There's the TIM which is the solder and it's quite low. And there's the chip itself which is somewhere in the middle of the two.

The new 9900K has a much thicker chip, and the solder is a thick layer. That means he saw a fairly impressive temp drop by getting rid of the thick solder for thin liquid metal. And he saw further gains by lapping down the CPU core. He did 0.2mm and said he could have gone deeper.

I might end up lapping my 9900K down by 0.3 or so mm because this upgrade is intended to be my last for a very long time, so if I can drop temps by 20 degrees then it'll make a lot less fan noise over the years.

tl;dr they'll all run hotter than 8700K would because there's much more material that's less conductive than there is in an 8700K. He's not quite sure but he's wondering if Intel had to make other compromises to get it soldered and as a result it's cancelled the gains.
 
The Halk said:
My understanding is a little different, the der8auer video is a good explanation of it and to be honest that's really the focus of his entire 9900k delidding video.

Basically there's 3 materials and each has a different level of conductivity. The more conductive the quicker heat will get through to whatever cooling you're using. There's the chip itself, there's the TIM and then there's the IHS. The IHS he explains is copper so it's high conductivity so that's not really the issue. There's the TIM which is the solder and it's quite low. And there's the chip itself which is somewhere in the middle of the two.

The new 9900K has a much thicker chip, and the solder is a thick layer. That means he saw a fairly impressive temp drop by getting rid of the thick solder for thin liquid metal. And he saw further gains by lapping down the CPU core. He did 0.2mm and said he could have gone deeper.

I might end up lapping my 9900K down by 0.3 or so mm because this upgrade is intended to be my last for a very long time, so if I can drop temps by 20 degrees then it'll make a lot less fan noise over the years.

tl;dr they'll all run hotter than 8700K would because there's much more material that's less conductive than there is in an 8700K. He's not quite sure but he's wondering if Intel had to make other compromises to get it soldered and as a result it's cancelled the gains.
Click to expand...
If you have the cpu already i would see what the performance and temps are like without delidding before deciding if you need to.

Not every 9900k is crazy hot. Some seem to be, but others seem right where they should be for the specs.
 
I don't have the CPU yet. I'm building a new build from scratch and I'm waiting to see how things settle for the CPU/MB/GPU thing. It's going to be quite an expensive build (4K plus I think)and also my last for several years.

As for the heat there is another issue to take into account. Everything I've said about the thicker die is correct, however there's _also_ an issue around power limits.

Issue number 2 is that it's rated as a 95W part and if it only gets 95W then it power throttles meaning it generates a lot less heat. If it gets all the power it needs then it'll run much hotter. Some of the reviews have been run where it looks very much like they're power throttling so reporting better thermals.

There is no way to get around though that compared to a delidded/liquid metal 8700K it will run much hotter, simply down to the thermal properties of the materials used.

https://www.overclock.net/forum/attachment.php?attachmentid=223640&d=1539304287

That's giving me a bit of pause for thought on the motherboard. It may be that overclocking these days when it's not simply chasing a high score number in ghz, when it's trying to increase performance of the parts will be quite different to how it was a few years ago. It might well be now that we do our best to feed the CPU/GPU power and take away heat and just let it do it's own thing.

I think all this will become clear within a few days.
 
tostitobandito said:
The new Maximus boards are 8 phase on the core (4x2 with doublers I believe), for 8+2 total.
Click to expand...

That is a 4 phase VRM though,so it is a (4+4)+2 phase VRM. There is still only 6 phases in total. This is nothing new - split phases have been used for years. However,just because a VRM has split phases does not mean it is any good,as it is dependent on the components used. A well engineered 4 phase VRM with sufficient cooling will be better than a split 4+4 phase with cheapo parts and insufficient cooling.
 
All the high-end Z390 (and Z370 for that matter) boards use doubler configurations though. By your logic, none have more than 8 actual VRM phases into a controller (EVGA Dark, MSI Godlike), and most are 4-6 phases. I believe this is because the voltage controllers themselves can't handle that many phases, so they double a smaller number of phases to get around that limitation. That said, everyone still refers to those two boards as 16 phases.

So that's why I'd call the Maximus boards 8 phase. I think the Z370 versions used essentially the same 4x2 VRM configuration, albeit with lower-spec chips.

Am I wrong?
 
tostitobandito said:
All the high-end Z390 (and Z370 for that matter) boards use doubler configurations though. By your logic, none have more than 8 actual VRM phases into a controller (EVGA Dark, MSI Godlike), and most are 4-6 phases. I believe this is because the voltage controllers themselves can't handle that many phases, so they double a smaller number of phases to get around that limitation. That said, everyone still refers to those two boards as 16 phases.

So that's why I'd call the Maximus boards 8 phase. I think the Z370 versions used essentially the same 4x2 VRM configuration, albeit with lower-spec chips.

Am I wrong?
Click to expand...
Hardware unboxed, very sure of himself, then not so much.

Hardware Unboxed15 hours ago (edited)
EDIT: I'll leave the original comment below but right now we're 99% sure the ROG Maximus XI Hero has a big fat 4-phase VRM designed to look like an 8-phase. So fake 8-phase that's really a 4-phase VRM, but we will continue to update this post as we get more info. ______________________________________________________________________________________________________ We've just been notified me that the ROG Maximus XI Hero does indeed feature doublers, so the 4-phases from the ASP1400BT controller are doubled, creating 8-phases. So, I’m not sure why it’s limited to 95 watts or why it was for Linus Tech Tips. We’ll keep digging into this, if the boards not limited by default it means the team over at Linus manually limited the board. That said Gamers Nexus states the following: "We have noticed that ASUS’ Maximus XI Hero we used follows Intel’s spec for boosting and power behaviour, unlike some other boards.”
Click to expand...

btw: phases listed >https://www.hardwareluxx.de/community/f12/lga-1151-mainboard-vrm-liste-1175784.html


Gamer Nexus @18:46 shows how to set the power limits.


From guru3d. Quick benches aren't seeing the clocks drop back. Unless disabled, could explain some of the results.

https://www.guru3d.com/articles_pages/intel_core_i9_9900k_processor_review,2.html
Again, as an example: the 9900K will boost to 4700 MHz on all cores. But after 30 seconds or something of full load that would drop to 4400 MHz. That last bit is highly dependant on the motherboard configuration though, often a turbo 'tweak' is applied so that the higher clocks will stay active. But a reference setup behaves as I just described. All processors we review are tested according to a reference setup.
Click to expand...
 
Last edited:
I've just run Cinebench on my 9900k and I'm seeing 2033. This was while downloading a few other Unigine benches for what it's worth.
I haven't changed anything in the ASUS BIOS except setting RAM to XMP profile 1 which is 3200 cl14.

I assume my CPU is boosting much higher than stock to get that number.
 
Doesn't anyone find it strange that 8Pack and der8auer tell gloom and doom about the 9900K heat problems.
And how you must get it delidded at a rip off price.

People do know they work for Caseking.de, who own ocuk?
 
deuse said:
Doesn't anyone find it strange that 8Pack and der8auer tell gloom and doom about the 9900K heat problems.
And how you must get it delidded at a rip off price.

People do know they work for Caseking.de, who own ocuk?
Click to expand...

Unless you have the very best air cooler or 240mm aio minimum, I'd say that's about right if wanting to overclock. As others have said, temps all over the place in reviews so it's difficult to know.
 
deuse said:
Doesn't anyone find it strange that 8Pack and der8auer tell gloom and doom about the 9900K heat problems.
And how you must get it delidded at a rip off price.

People do know they work for Caseking.de, who own ocuk?
Click to expand...

I also think its fair to acknowledged that even though Intel have soldered these, they have done so poorly and that by de-lidding and lapping there are further reductions in temperature which for those who do want to overclock and want best possible results, any reduction in temperature is welcomed.

End of the day there is no gun to your head, just buy the regular CPU, if you want an 8 Pack or der8auer version, then you pays the price, simples. :)
 
Drollic said:
Unless you have the very best air cooler or 240mm aio minimum, I'd say that's about right if wanting to overclock. As others have said, temps all over the place in reviews so it's difficult to know.
Click to expand...


It's better to buy a AIO\Top air cooler than to have it delidded.

And der8auer only got 13c less after delidding and lapping a RETAIL 9900K
 
Last edited:
Other results for delidding several CPUs put the temp change at 4-7 degrees for the delid, with an additional 1-2 degrees from sanding the die down an extra 0.2mm.
Using liquid metal after delid ofc.
 
You must log in or register to reply here.
Back
Top Bottom