[H] delided a Kaby Lake i7 7700K, the results are .. interesting

[Notional]

This is why Ryzen is so important. Intel is like a fat lazy greedy corporation. IPC goes nowhere. Core count goes nowhere. Prices rise slowly and constantly. It's about time we get some proper competition. Of course, that doesn't happen if people still buy Intel per automatics, even if it's much worse in price to performance.

[Lays]

7 minutes ago, Lord Nicoll said:

No they don't I know a few who uses bare die. The risk of shorting is very small as long as you keep the pot below ambient all the time, and insulate correctly, bare die gives much better temperatures.  

Nick Shih is one of the only people that has done it recently when he validated a 6700k at like 7+ g, everyone else is cpu->grizzly->ihs->grizzly->pot or gc extreme. 

 

Mass and surface area are EXTREMELY important at -190c, at low load it's OK to do no IHS, but any long benchmark it's not recommended. The extra mass that the IHS adds helps with temp control if the CPU has a cold bug, and increases surface area.

[Lord Nicoll]

20 minutes ago, Lays said:

Nick Shih is one of the only people that has done it recently when he validated a 6700k at like 7+ g, everyone else is cpu->grizzly->ihs->grizzly->pot or gc extreme. 

 

Mass and surface area are EXTREMELY important at -190c, at low load it's OK to do no IHS, but any long benchmark it's not recommended. The extra mass that the IHS adds helps with temp control if the CPU has a cold bug, and increases surface area.

When you've a 2 kilogramme LN2 pot, it isn't that important, The IHS if reapplied badly can create a trap zone for moisture. 

[Jumper118]

3 minutes ago, Lord Nicoll said:

When you've a 2 kilogramme LN2 pot, it isn't that important, The IHS if reapplied badly can create a trap zone for moisture. 

yeah but its really easy to crush the die with a 2kg pot + the huge mounting pressure. i would never bench naked die on skylake or kaby as the pcb is too thin. 

[Kobathor]

10 hours ago, Citadelen said:

I'm pretty sure the GTX 480 is even hotter, get four of them

i don't want my house to burn down

[Lord Nicoll]

1 minute ago, Jumper118 said:

yeah but its really easy to crush the die with a 2kg pot + the huge mounting pressure. i would never bench naked die on skylake or kaby as the pcb is too thin. 

That's what the delid bracket is for, and either way, an IHS sits on the die either way, whether it's there or not doesn't matter, it just makes the normal mounting bracket mount it better. 

[Jumper118]

1 minute ago, Lord Nicoll said:

That's what the delid bracket is for, and either way, an IHS sits on the die either way, whether it's there or not doesn't matter, it just makes the normal mounting bracket mount it better. 

but only msi boards have that and msi boards are crap  

[Lays]

1 minute ago, Lord Nicoll said:

That's what the delid bracket is for, and either way, an IHS sits on the die either way, whether it's there or not doesn't matter, it just makes the normal mounting bracket mount it better. 

Which people do you know using bare die? I've been competing on hwbot for 2-ish years now and talk to a lot of the top guys, Nick Shih is literally the only guy I've seen in the past year doing it with a 7g+ validation on a 6700k.

[Lord Nicoll]

1 minute ago, Jumper118 said:

but only msi boards have that and msi boards are crap  

I know MSI make one, but there are much better ones and they're universal . I also dislike MSI boards though, ASRock too, but they usually cost less so win win. 

[Lord Nicoll]

5 minutes ago, Lays said:

Which people do you know using bare die? I've been competing on hwbot for 2-ish years now and talk to a lot of the top guys, Nick Shih is literally the only guy I've seen in the past year doing it with a 7g+ validation on a 6700k.

Buildzoid for one does (did, idk, haven't watched him lately) it. I only did LN2 with a friend once, I can't say I do it, but I like to run my hardware 24/7 for HPC so I tend to actually underclock. 

[flibberdipper]

Tech City just did a video on this, and the results of his 7700K with Liquid Pro was fucking astonishing.

 

 

[Lays]

1 minute ago, Lord Nicoll said:

Buildzoid for one does (did, idk, haven't watched him lately) it. I only did LN2 with a friend once, I can't say I do it, but I like to run my hardware 24/7 for HPC so I tend to actually underclock. 

Buildzoid doesn't even have a kaby or skylake cpu dude. LOL

[Lord Nicoll]

1 minute ago, Lays said:

Buildzoid doesn't even have a kaby or skylake cpu dude. LOL

Yeah, he did it on Devils Canyon, there isn't a difference beyong 14nm and a slightly thinner PCB (which doesn't matter if you mount it right, but looking at some LN2 rigs, overclockers rarely do things right), it works the same. 

[Lays]

Just now, Lord Nicoll said:

Yeah, he did it on Devils Canyon, there isn't a difference beyong 14nm and a slightly thinner PCB (which doesn't matter if you mount it right, but looking at some LN2 rigs, overclockers rarely do things right), it works the same. 

@Jumper118 just asked him on Telegram if he's ever benched naked die on LN2, and he said nope.

[Lord Nicoll]

2 minutes ago, Lays said:

@Jumper118 just asked him on Telegram if he's ever benched naked die on LN2, and he said nope.

Ok, and? He's one, just because some don't doesn't all don't. It's probably easier with it, but that's not what extreme overclocking is about. The IHS has often been left off, it was more popular in the past, but in the end you're still having an extra TIM layer and a jenky mount between them. 

[vanished]

2 hours ago, patrickjp93 said:

Reduction in Temperature =/= Reduction in Heat/Thermal Energy. The statement is correct.

Let me requote the original statement:

 

Quote

Replacing our Kaby Lake TIM resulted in a 25.28% decrease in package temperature

 

OK, relative to what?  If it dropped to 0 C, does that mean there was a 100% reduction in temperature?  What about -20 C?  This is meaningless and nonsensical.  I could understand them perhaps making it relative to the idle temps, but this is totally arbitrary and it's a stupid comment. 

 

6 hours ago, FPSwithaWacomTablet said:

Q = mcT

 

Where Q is the energy of the object in Joules, m being the mass of the object, c being the heat capacity for a given unit mass of said object and T being the temperature in Kelvin.

 

It's a quick tour of high school physics guys. Though if you really wanna dig down deep into temps you might as well start learning thermodynamics.

 

TL;DR It just means it's easier to make light things hotter than heavy things, all else being equal.

 

Should be dQ = mc(dt) but can't be bothered to find the delta symbols and calculus

Yes, but I'm not sure how that comes into play here.

5 hours ago, Devin92 said:

I dont really understand this. However my knowledge is still from high school so correct/update me. From my understanding of specific heat capacity, temperature change and the amount of energy needed to absorb (dissipate) has a linear ratio, given the specific heat does not change with temperature. so a 20C chip do has double the amount of energy compares to a 40C chip.

No, the problem there is that 0 is a totally arbitrary number.  If you convert to any absolute scale, like Kelvin for example, you can then say that 293.15 K (20 C) is "twice as hot" as something at 146.575 K (-126.575 C) but you can't do that in Celsius.  40 C (313.15 K) is only "6.8% hotter" than something at 20 C (293.15 K), if we are going to use that language, which I really would rather not.  It makes a lot more sense to say something at 40 C has 6.8% more thermal energy than something at 20 C (assuming a constant specific heat capacity and unchanged mas sand composition).

[Devin92]

2 minutes ago, Ryan_Vickers said:

No, the problem there is that 0 is a totally arbitrary number.  If you convert to any absolute scale, like Kelvin for example, you can then say that 293.15 K (20 C) is "twice as hot" as something at 146.575 K (-126.575 C) but you can't do that in Celsius.  40 C (313.15 K) is only "6.8% hotter" than something at 20 C (293.15 K), if we are going to use that language, which I really would rather not.  It makes a lot more sense to say something at 40 C has 6.8% more thermal energy than something at 20 C (assuming a constant specific heat capacity and unchanged mas sand composition).

ya. you are right. damn, forget about the unit of temperature. dumb by me.

[VerticalDiscussions]

12 hours ago, Ryan_Vickers said:

-snip-

.-.?

[vanished]

1 minute ago, VerticalDiscussions said:

.-.?

I tried to elaborate just a few posts back, go to page 4  

[VerticalDiscussions]

3 minutes ago, Ryan_Vickers said:

-snip-

 

You mean my life is a lie o-o? RIP

 

 

20ºC is not x2 40ºC and i.... i.... and then its the same for Kelvin but you say its a bad scale to use and its 6.8% and... and that the size and shape of the body affects how much higher the temperature is relative to the maths and....

 

 

 

 

[vanished]

1 minute ago, VerticalDiscussions said:

You mean my life is a lie o-o? RIP

 

 

20ºC is not x2 40ºC and i.... i.... and then its the same for Kelvin but you say its a bad scale to use and its 6.8% and... and that the size and shape of the body affects how much higher the temperature is relative to the maths and....

Unless I've made some horrible error without realizing it, I think you might have misinterpreted it  

40 = 2x 20 does not work in C because 0 is an arbitrary value with not meaning in this context.  With an absolute scale like Kelvin, you can simply multiply a number by two (ie, 200 K = 2x 100 K) because 0 is absolute zero, where there is no heat at all.  It's a good scale for that reason...

[VerticalDiscussions]

2 minutes ago, Ryan_Vickers said:

-snip-

So Celsius sux hairy balls :D?

 

 

[WkdPaul]

4 minutes ago, Ryan_Vickers said:

Unless I've made some horrible error without realizing it, I think you might have misinterpreted it  

40 = 2x 20 does not work in C because 0 is an arbitrary value with not meaning in this context.  With an absolute scale like Kelvin, you can simply multiply a number by two (ie, 200 K = 2x 100 K) because 0 is absolute zero, where there is no heat at all.  It's a good scale for that reason...

pretty much yes, basically, if you changed scale (let's say fahrenheit), then the 25.?% they mentinoned doesn't work anymore.

 

Your explanation on last page was spot on BTW!

Edited February 3, 2017 by wkdpaul
typos, words be difficult

[AresKrieger]

5 hours ago, Devin92 said:

I dont really understand this. However my knowledge is still from high school so correct/update me. From my understanding of specific heat capacity, temperature change and the amount of energy needed to absorb (dissipate) has a linear ratio, given the specific heat does not change with temperature. so a 20C chip do has double the amount of energy compares to a 40C chip.

If you use kelvin then yes however this is is off set by 273.15 so its going from 293.15 to 313.15 thus about a 7.5% improvement if you went from 40C to 20C

[vanished]

3 minutes ago, VerticalDiscussions said:

So Celsius sux hairy balls :D?

 

 

I wouldn't say that.  It's a lot better for scientific purposes than something like Fahrenheit imo but you will see Kelvin used more commonly still for the reason I mentioned.  Of course, in certain equations there's no difference (ie, if you're taking the difference between two temperatures), but generally you want to be consistent to be safe.