Does Haswell use crappy TIM between the IHS and silicon?

KazeoHin

[H]F Junkie
Joined
Sep 7, 2011
Messages
8,992
Hey guys, can anyone confirm if the new i5 and i7 4XXX series chips use cheap TIM between the IHS and the silicon? [H]ardOCP's review hinted that it was (unfortunately) using the same IHS setup as Ivy bridge CPUs, instead of running the same setup as Sandy Bridge (soldered contact).

Is anyone brave enough to attempt a De-Lid?
 
Why did they use the same method if it was clear it didn't work successfully last time?
 
Why did they use the same method if it was clear it didn't work successfully last time?

Because it is cheap and they aren't concerned with overclockers.

If you want better TIM and just overall better product go Sandy Bridge (if you can find them) or Sandy Bridge E.
 
Makes more sense, I assumed it would translate and be just as poorly performing at the stock temperatures as well. Will take a peak at Sandy-E but more likely going to go for a 4.2-4.4gHz Haswell OC.
 
It's not that damn hard to de lid the bugger, just do it careful I've always done this and it does help alot.
 
If you want better TIM and just overall better product go Sandy Bridge (if you can find them) or Sandy Bridge E.

What do you mean if you can find Sandy Bridge? It's in stock at newegg. I'm sure if newegg has them they are findable no? I've never seen newegg sell stuff that's super rare before.
 
Plus, would a 4.7ghz Sandy compete with a 4.4ghz Haswell? Maybe the extra OC leverage isn't AS important if the performance isn't there. I know sandy's can OC to 5+ with certain chips, but by the same token, Haswells can go to 4.7 with the right chip. So how would their performance compare?
 
Plus, would a 4.7ghz Sandy compete with a 4.4ghz Haswell? Maybe the extra OC leverage isn't AS important if the performance isn't there. I know sandy's can OC to 5+ with certain chips, but by the same token, Haswells can go to 4.7 with the right chip. So how would their performance compare?

about equal. The IPC difference is minimal, and is more or less made up for with the nominal OC advantage of the older chips.

It's a wash.
 
about equal. The IPC difference is minimal, and is more or less made up for with the nominal OC advantage of the older chips.

It's a wash.

And there's potential in the future for Haswell to be faster, with support for AVX2 and FMA. Haswell's IPC advantage is just running older code.

This is very important now that we're on 4-5 year upgrade cycles for CPUs. More and more programs that need extra compute power will be optimized for new instructions.
 
Plus, would a 4.7ghz Sandy compete with a 4.4ghz Haswell? Maybe the extra OC leverage isn't AS important if the performance isn't there. I know sandy's can OC to 5+ with certain chips, but by the same token, Haswells can go to 4.7 with the right chip. So how would their performance compare?

The right chip, but the % of chips that overclock high for Haswell aren't nearly as high as they were for Sandy. Plus you will have to invest in more expensive cooling for Haswell than you would Sandy.
 
The right chip, but the % of chips that overclock high for Haswell aren't nearly as high as they were for Sandy.

You can't possibly form this conclusion yet.

Plus you will have to invest in more expensive cooling for Haswell than you would Sandy.

Clock for clock, yes, but performance for performance no.
 
I love the delidded hold down bracket that the m6e is going to come with
 
Most of the guys and gals that are serious about over clocking prefer the ability of haswell to be delidded for bare die cooling.

Thank you Intel.
 
Most of the guys and gals that are serious about over clocking prefer the ability of haswell to be delidded for bare die cooling.

Thank you Intel.

So Intel is paying people to post in the Hardforum now?
 
I can't believe they did the same screw up (in my opinion it's a screw up) as IB.
With all the complaints with IB you think with their new "baby" they would do it right.
 
Did they ever improve on it after release for IB? Wondering if I should forgo early adoption in hopes a more refined final product will hit in the next couple months.
 
I was thinking. Would it be easier to delid Haswell or IB by sticking in the freezer and then taking a stab at the lid with the razor? (no pun intended)
 
It looks like people have problems with daggers. You don't take a stab, you gently and slowly delid your CPU.
 
I'm thinking this is just 'how it will be' from now on. I wouldn't be surprised if the new enthusiast level chips are manufactured using generic-brand TIM instead of solder. Intel is simply not listening to the hardcore crowd. At least some complaints get an acknowledgment like 'deal with it' but intel probably has NO IDEA that this is an issue. Now pretty much every processor from now on will be manufactured this way: end of an era.
 
If only intel sold these chips without a lid.

it's very easy to remove the lid with hammer/vice method.

For the majority of users, the IHS doesn't impede their use much... and it's much easier than clamping the HSF down onto the bare die.
 
[CoF] ATOMIC;1039936908 said:
I was thinking. Would it be easier to delid Haswell or IB by sticking in the freezer and then taking a stab at the lid with the razor? (no pun intended)

huh? are you planning on the freezing loosening the TIM?
 
Hey guys, can anyone confirm if the new i5 and i7 4XXX series chips use cheap TIM between the IHS and the silicon? [H]ardOCP's review hinted that it was (unfortunately) using the same IHS setup as Ivy bridge CPUs, instead of running the same setup as Sandy Bridge (soldered contact).

Is anyone brave enough to attempt a De-Lid?

http://www.overclockers.com/3step-guide-to-overclock-intel-haswell

Speaking of temperatures, surely people will be saying “Haswell has the same problem as Ivy Bridge,” referring to the thermal paste issue. Yes, Haswell has thermal paste, but from Very Authoritative People, the TIM is not the problem. As has been posited across the net when people de-lid Ivy Bridge chips with great results, it’s really the black adhesive that’s the culprit. When you cut out that adhesive, it allows the IHS to sit closer to the CPU die, meaning there is less thermal paste through which the heat has to travel, leading to significantly lower temperatures. Intel’s TIM is really quite good, but the manufacturing process leads to that glue being just a little too thick, which is why you see such temperatures.
 
Hmm, makes me wonder if scraping the black glue off the IHS and allow it to sit closer to the die would help out a ton if one were to replace the TIM with something like CL Ultra.

Die -> CL Ultra -> (lowered) IHS -> CL Ultra -> HSF

Thoughts?
 
Hmm, makes me wonder if scraping the black glue off the IHS and allow it to sit closer to the die would help out a ton if one were to replace the TIM with something like CL Ultra.

Die -> CL Ultra -> (lowered) IHS -> CL Ultra -> HSF

Thoughts?

I can't comment on the glue part, since I removed it once I delidded my chip, but I guess it couldn't hurt.

Another thing you could try, is directly cooling the die.
 
too much risk for such a expensive processor, id be damned if the process damaged the processor that would few years worth of upgrade money down the drain for me :(
 
Back
Top