Haswell delidding, any news?

owell... it was fun, and went pretty smooth. Started in one corner, and worked my way around... only took ~15min. im ~15 degrees cooler with higher clocks under load, and get to notch my belt!!!!!!!!!!!!!!

I guess everyone's view on delidding differs, which is cool.

I am one of those people that don't mind my chip 'naked'. I mean, I pressue a large metal water block on my GPU on PK-1, why not the CPU too?

I am however very careful about 'not crushing' the chip, I bring lab equipment to measure everything before pulling the trigger. So far, I have to say, I think delidding.. not a bad thing to do, but for those who never did it before, and might not have the awareness of what 'can' go wrong, do think twice.

back to haswell.. I am a fair bit disappointed they used crappy TIM still for the chip.. sigh.. I thought the consumer's voice there was loud enough from the IB crappy TIM incident... apparently not loud enough..

Intel probably sees it as "good enough, and just the right price"

Tim and gap causing 20C increase in temp at higher volts is same issue.

IHS to die requires a larger gap (than IHS/waterblock) to decrease voids/pump out from thermal cycling. STIM (solder) fills the bill nicely since intels variant has a thermal conductance of 80 W/mK, about 15X higher thermal conductance than best paste TIM, 150 microns of solder is not a thermal barricade, like 150 microns of paste. Most pastes are tested to be used at 20-30 microns thickness.

Problem is at stock watts, stock clocks, newer cpus IVY/haswell falls below the TDP for which intel developed solder via multiple white papers, despite die shrink increasing thermal density.

When overclocking, the dissipated power would hit the point intel would use solder, but intel doesnt design specs for OCing, that is bonus only. At stock, only 3-5C difference, isnt worth extra cost for intel, operates well within margins, even at moderate overclocks.

Only when using high volts and getting last 200mhz, does delidding help.

I would delid just for fun, but wouldnt use it for 24/7, only because of pump out, having to reapply, and general pita when changing out hardware, new cooling equipment, etc.

No pump out with liquid metal tim...other than that, you seem to be spot on. Intel will never see this a problem.

yep, those using liquid metal dont seem to have pumping out issue of those that tried tim...the liquid ultra causes a bonding problem with waterblock and IHS, but when using IHS and die that problem would be beneficial. Even pro havent seen people complaining of having to reapply, unlike those replacing with paste and seeing decline over time.

If I delidded, I would use liquid ultra and hope it forms a bond and pseudo permanent attachment.

btw, someone already delidded 4770K, at 1.225v core, saw 11C drop in max temps, looks similar to IVY bridge so far....here

I was thinking the USB 3.0 sleep bug was on mobo, but forgot it is actually the cpu that has USB 3.0 and bug, and will be fixed in future ?silent stepping.

So may get 4770k, dont mind changing out cpu in few months which takes little time, mobo to big pita to change in few months.

I have to decide if want to play with new toy or not....I could even delid the USB bugged one since will replace it in time anyways...

rge said:

yep, those using liquid metal dont seem to have pumping out issue of those that tried tim...the liquid ultra causes a bonding problem with waterblock and IHS, but when using IHS and die that problem would be beneficial. Even pro havent seen people complaining of having to reapply, unlike those replacing with paste and seeing decline over time.

If I delidded, I would use liquid ultra and hope it forms a bond and pseudo permanent attachment.

btw, someone already delidded 4770K, at 1.225v core, saw 11C drop in max temps, looks similar to IVY bridge so far....here

I was thinking the USB 3.0 sleep bug was on mobo, but forgot it is actually the cpu that has USB 3.0 and bug, and will be fixed in future ?silent stepping.

So may get 4770k, dont mind changing out cpu in few months which takes little time, mobo to big pita to change in few months.

I have to decide if want to play with new toy or not....I could even delid the USB bugged one since will replace it in time anyways...

Click to expand...

I believe the usb issue is part of the chipset and Intel is already shipping revised steppings

yeah, that is what I thought originally....then read something confusing, implying cpu. But article quote below agrees it is just chipset, and quote from intel.

http://us.hardware.info/news/34321/...0-bug-in-haswell-updated-with-intel-statement

4th gen Core is on track for a mid-year launch. Intel issued a PCN documenting a chipset USB errata and stating that chipsets with the errata will be in production during the initial ramp. But Intel has confirmed that there is no chance of data loss or corruption. This issue has only been observed with a small subset of USB SuperSpeed thumb drives and does not affect other USB peripherals. We take all customer issues seriously and should any customer have a question or concern they can always contact Intel customer support.”

Click to expand...

Even though mobo manufacturers already should have it...says in one article, they wont be shipping those with USB fix until july 15, dont know if that quoted date from article is reliable or not.

Intel will start to provide samples of 8-series/C220-series chipsets C2 stepping to customers starting from April 19, 2013. The first shipments of the fixed chipsets will begin on July 15, 2013.

Click to expand...

link below you can see a chart with affected mobo chipsets:
http://us.hardware.info/news/34321/...0-bug-in-haswell-updated-with-intel-statement

anyone with haswell mobo can see if they have C2 stepping chipsets per graph in link above. But if only affects some thumb drives and not other USB, wont be too bad.

I just ordered myself sont Luquid Ultra and Pro "TIM", will jump onto delidding very soon

I delid my 4770k successfully yesterday using the hammer/vice method. Was much easier than I expected and temps dropped by about 20deg. Totally worth it.





4.9Ghz @ 1.47v:


4.6Ghz @ 1.28v only gets about ~48deg peak which is what I run 24/7

nice one. I will be a little anxious to do the hammer/vice thing.. I much rather slowly use the razor to carefully take off the IHS. Maybe I am chicken. bwak bwak bwaaaquak

I'm currently testing alot of paste combo on die/ihs and ihs/heatsink. Stay tuned for my results

bluezero5 said:

nice one. I will be a little anxious to do the hammer/vice thing.. I much rather slowly use the razor to carefully take off the IHS. Maybe I am chicken. bwak bwak bwaaaquak

Click to expand...

The razor method seems much scarier to me, just a small slip and you damage the die. The vice/hammer method was really easy and quick - you'd have to be really un-co to screw it up.

I think we all have our own "scares" about how to delid a CPU. Some prefer the vise and hammer, other prefer blades....

I prefer blades. And you ?

Fair enough, I was completely nervous before I got the ol' hammer out, but after doing it and seeing how quick and easy it was (took me like 30 secs), I can't help but recommend it.

bluezero5 said:

back to haswell.. I am a fair bit disappointed they used crappy TIM still for the chip.. sigh.. I thought the consumer's voice there was loud enough from the IB crappy TIM incident... apparently not loud enough..

Click to expand...

Also, this

Jesus, 4.9 on haswel on air, almost makes me want to upgrade, i feel like i am missing out on all this delidding fun!

i want a haswell @ 5ghz now *grumble grumble*!

I keep bouncing back between deliding and not. everytime I read an arguement for the other side my mind says "hmmm good point better do it that way" lol

Lt_Horn said:

I keep bouncing back between deliding and not. everytime I read an arguement for the other side my mind says "hmmm good point better do it that way" lol

Click to expand...

The best point to delidding is when your overclock is limited by temperature, not by voltage.

If you're hitting a voltage wall then delidding won't help you, but if you're at say 1.1V and hitting 95C, then it'll help.

Well, I just tried the hammer and vice method, and everything went smoothly. It reduced my temps by about 20c.