• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Intel still using inferior thermal interface?

Haswells are particularly victim of noobshield implemtation. I've heard of people seeing 20 degree drops but that is baredie. Someone found that its not so much the Tim with haswell as it is a misshapen underside of the noobshield. IIOW dont bother lapping anything. There is a lot less clearance to cut with a blade since Intel chips have come with the lip around it so it is a lot easier to cut a trace. A method that is gaining popularity is putting the chip in a table vice and tapping a block of wood against the PCB. many videos on YouTube about it. All my CPUs are nekkid its the only way to overclock effectively.
 
I would never do that unless I got a particularly hot chip, or I was trying for a MAX oc.

Some people do it just because they like to tinker....but to do it for an arbitrary temp goal? meh...
 
Temperature drops in excess of double digits are not arbitrary. A general rule of thumb is that for every ten degree celsius drop, an ICC's life is doubled. These noobshields have been noisome ever since their reintroduction as far as cooling capacity goes.
 
PliotronX said:
Temperature drops in excess of double digits are not arbitrary. A general rule of thumb is that for every ten degree celsius drop, an ICC's life is doubled. These noobshields have been noisome ever since their reintroduction as far as cooling capacity goes.
Click to expand...

I haven't ever heard of problems with CPUs failing much. All of these people complaining about WinXP support are probably on P4 dinosaurs, like much of the workforce.

Being under warranty in the first 3 years is more important to me than possibly having a working legacy chip in 25 years.......but to each their own. 😛
 
Would there be any danger in direct die contact like back in the athlon XP days. They have small dies, like the Barton core, and if you were not an idiot, they would be just fine.

The new 4 corner mounting system on 1150 is superior, so I would think it would be safer overall.
 
AMD uses the same "inferiour" paste TIM on all FMx CPUs.

Trinity and Richland delidded.

LL

500x1000px-LL-c97faa64_iso_ihs_die.jpeg
 
Homeles said:
The TIM Intel uses is actually quite good; it's just not as good as solder.
Click to expand...

The TIM isn't the issue, it's the imprecision of the application of that black sealant. Some its fine, others it creates an air gap between the TIM and the IHS.

People shouldn't be able to realize 20oF gains by switching from normal HSFs to ones with high clamping force like NH-D14.
 
OCGuy said:
I haven't ever heard of problems with CPUs failing much. All of these people complaining about WinXP support are probably on P4 dinosaurs, like much of the workforce.

Being under warranty in the first 3 years is more important to me than possibly having a working legacy chip in 25 years.......but to each their own. 😛
Click to expand...
Stock chips can take a beating but if operating beyond specs which most of us geeks do the cooler temps result in being able to run higher clocks at lower voltage. Even my Q6600 was able to run higher with lower vcore and vtt than with a hand lapped noobshield and TRUE to 2000 grit.
 
Fjodor2001 said:
Why use TIM instead of solder? Is it to save a few cents per chip, or what's the reason?
Click to expand...

We'll never know, although is has posited that it is to due with thermal expansion and flexing. Maybe solder couldn't hold up to the expansion and contraction rates of the differing materials that make up the CPU package.
 
Last edited:
Phynaz said:
We'll never know, although is has posited that it is to due with thermal expansion and flexing. Maybe solder couldn't hold up to the expansion and contraction rates of the differing materials that make up the CPU package.
Click to expand...

You would think if that were the case that they couldn't solder HW-E either.
 
So here is a concern. I am going to be building a haswell system. I just ran prime on a dell optiplex 3020 with haswell i5-4570. At 100% cpu, the core temps sky rocketed up to 85c. Is this safe? It is hitting thermal throttling yet?

In comparison my optiplex 790 hit 74c w/ i5-2400 sandy.

What gives. These are bone stock.
 
Phynaz said:
We'll never know, although is has posited that it is to due with thermal expansion and flexing. Maybe solder couldn't hold up to the expansion and contraction rates of the differing materials that make up the CPU package.
Click to expand...
The 22nm process had issues early on with "thermo-mechanical/moisture." It's entirely possible. It's also entirely possible that it was a money-saving move. It could even be a bit of column A and B.

http://www.semiwiki.com/forum/files/Intel Manufacturing Slides 2013.pdf ~slide 4
 
Fjodor2001 said:
Why use TIM instead of solder? Is it to save a few cents per chip, or what's the reason?
Click to expand...

Originally, if you read through the "War and Peace" of IDontCare's groundbreaking "De-lidding experiments" thread, it was assumed by some that Intel was afraid of damaging the chip with the Indium solder because of its its lithography -- shrunk to 22nm.

But all the "E" processors including Sandy Bridge and Ivy-Bridge -- especially Ivy Bridge -- use solder when the socket-1155 Ivy-Bridge cores use the TIM. As someone said, this in turn made for a less accurate or precise fabrication process, so there was likely wider variation in the distance between the IHS and the die than with Indium solder.

But I think it ultimately became apparent that Intel did this for reasons of cost: they had reduced the TDP spec for the newer processors, and decided to reduce production costs as a result. Otherwise, there would not have been a choice to use Indium solder with the IB-E processors, which have more cores and the same lithography.

I (or we) expect that Intel will use Indium solder with the Haswell-E processor, versus the TIM they used for the Socket 1150 units.

If you wanna be an enthusiast, you either pay for the processor, or pay with your time, sweat and risk-generated tears. Otherwise, suck it up. They made these processors primarily for a market of folks who buy OEM machines and don't care about over-clocking or other abnormal compulsive obsessions.

The rest of us -- a minority -- are "S--O--L." If you don't know what that means, I'm sure you can look it up.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top