Intel Core i7-5960X Haswell-E CPU Question

[cdawall]

Or the fact that no amd even the Hugh watt chips have one. Some are soldered some are not. It's really interesting the entire thing.

 

[RealNeil]

Wouldn't applied thermal interface material along with the heatsink's contact with the CPU cover that hole?

 

[lilhasselhoffer]

Wouldn't applied thermal interface material along with the heatsink's contact with the CPU cover that hole?

A point I made much earlier.

Again, we come to the fact that there seems to be no hard and fast rule. It isn't all epoxied chips. It isn't all soldered chips. All that remains is Intel basically just adding a hole whenever the production run has issues, and they have to add the extra hole to alleviate...something. It's less than a satisfactory answer in my book, but it seems like the best we'll get. Dang.



Edit:
Here's the earlier quote:

...
We've established why it just being a pressure release is rather unlikely, without even asking whether it would actually breathe. I mean, the IHS should be covered by both a heatsink surface, and thermal paste. The implication would be that the pressure differential between the CPU air and surrounding air would be enough to actually blow the thermal paste out of the way, right? I'm not sure how this is possible, after you stated clearly that you needed to clean the material out of the holes.
...

I believe this is the same point you raised, but if not please correct me.

 

[EarthDog]

Let me see what I can find...

 

[Potatoking]

normal haswell has hole in the black glue holding the IHS haswell-e has glue all around
maybe?

 

[cdawall]

Someone PM w1z see what he has to say?

 

[natr0n]

Its for gasses to escape the IHS when manufacturing.

 

[EarthDog]

Its for gasses to escape the IHS when manufacturing.

If only it were that easy... see previous posts.

 

[cdawall]

If only it were that easy... see previous posts.

I mean it could still be they could be using a different process for different processors.

 

[EarthDog]

We have seen both methods/TIMs with holes and without, righT?

 

[revin]

Just a stab but it may be for hanging for passivation.
We had to do that for lead parts that we glued to steel parts for airport X-Ray equipment.
It was more consistent to hang than just "bulk" dip with a basket, or as with the larger items by hand.
The only "requirement" on the print was "Passivation to all part's before gluing to ensure clean surface"
Albeit we would have to get an engineering approval before hand to put in hole's

 

[cdawall]

We have seen both methods/TIMs with holes and without, righT?

We don't know the exact methods and tims used. Just because one is soldered does not mean it used the seem solder and process as the other. Same goes for the epoxy style.

and holy hell Revin your signature quote is old as fuck

 

[EarthDog]

Let me see what I can find...

Not much. I asked a buddy that worked at Intel who then asked some engineers... not the right ones apparently. LOL

Since we have some Intel reps in the lab atm, I was able to ask them directly. None of the engineers know! lol. Their best guess is that its used for alignment, spreading out paste, or allowing air flow of some sort through the chip.

Sorry I couldn't help further. Since they don't know, I would bet its a design secret. Intel is sneak like that, and a lot of teams inside of Intel are not aware of why a team will perform a task.

 

[cdawall]

@W1zzard what is your take on all of this?

 

[W1zzard]

@W1zzard what is your take on all of this?

I never heard any explanation other than for thermal expansion of air under the IHS

Isn't that the equivalent on Sandy Bridge? (The red rectangle, ignore the red circular marker, I just used Google images)

 

[lilhasselhoffer]

Then I eat my words. Apparently my logic jumped the track, when I assumed that the hole was the only way you'd account for an opening in the die. Much obliged, and my apologies for dragging this discussion out for much longer than necessary.