What holds you back from running your delided cpu neked?

[caddi daddi]

I think I run 7 delided cpus neked, I have no issues at all doing this, they all run just fine.
I can't understand taking the risk (very little in my book) of deliding the little lump and just reinstalling the offending IHS, what is the fear?
let me count up what I run buck neked.
4690K
8600K
3, 4790K
3, 8086K
6800K
ok, that comes out to 9 that I run 24/7, buck friggin neked.
why do you reinstall the IHS?

 

[EarthDog]

I tighten up like a clam when I hear naked CPUs.

To me, the risks outweigh any benefits.

 

[trents]

I have killed two i7s and a motherboard or two trying to go naked. It happened once when I even bought one of those shem kits that was supposed to prevent that. If ED tightens up like a clam I tremble like a wet dog on a cold day. Besides, from what I read it does little or nothing to drop temps compared to delidding. And I have delidded numbers of times without incident. I think LGA sockets make going naked a risky proposition. They are so fragile. If you don't get the pressure just right and evenly distributed you risk damaging something. I remember years ago AMD CPUs came naked from the factory with little rubber support pads on the PCB to give the right clearance to the heat sink and distribute the pressure. I never had any problem with those. But they were PGA, not LGA socket chips.

 

[ShrimpBrime]

I think I run 7 delided cpus neked, I have no issues at all doing this, they all run just fine.
I can't understand taking the risk (very little in my book) of deliding the little lump and just reinstalling the offending IHS, what is the fear?
let me count up what I run buck neked.
4690K
8600K
3, 4790K
3, 8086K
6800K
ok, that comes out to 9 that I run 24/7, buck friggin neked.
why do you reinstall the IHS?

I've run 14 soldered de-lidded AMD processors naked. Lots of athlons and semprons from s462 to AM3.

Knowing this is not related directly to de-lidding of Intel processors, from what I do know, AMD is easier to deal with. I was enlightened recently that Intel chips have much less surface area and seem to be fragile if too much pressure is applied. And typically with de-lidded processors, you need a decent amount of clamping force. So the Cpu PCB and or motherboard PCB can bend, which I've seen cause no post issues with my higher TDP AMD processors.

The nice thing about removing layers of copper and thermal paste is you have a much faster transfer of temperature from Cpu core to waterblock. However it seems to have the biggest impact running a much higher TDP than most people would do. I found in most cases an average of about 10c can be lost with running naked. My FX-5000 didn't seem to make much difference as it didn't clock well having the TLB errata issues only reached 2750mhz unlocked. could have done that on an air cooler really.....

I do have a 2700X that some time in the future (no actual deadline set) to de-lid. For now I'm still getting used to the processor and currently running the stock air cooler and heavily limited by temperature near the 1.5v area which I need to surpass the 4.4ghz barrier I'm at now. But then again, it's plenty fast, I run it daily at 4.2ghz typically and leave the voltage on auto believe it or not....

Any Hoot, here's that soldered AMD naked list I've accumulated through the years, just in case some one thinks I've not been down that naked road once or twice. (Have done a couple of Intel chips also )
Here's 14 working cpus and I did Kill 2 so 16 total. Cannot give a direct number on non-soldered processor, we'll say maybe a few dozen or so?

FX-4300
FX-6100
FX-9590
FX-8320
B73 x3
Athon 5000+ unlocker FX-5000
970T
1090T
965BE
940BE
9850BE
9950BE
6400+ x2

 

[maxfly]

i havent run anything nekkid since the socket 478 days. i had an asus p4c800-e deluxe mb that i bought a asus ct-479 pentium m socket adapter for.
https://techreport.com/review/8585/asus-ct-479-pentium-m-socket-adapter
it allowed me to run bare mobile procs and yeah i killed more than one by trying to modify the weak *** cooler to accept water blocks and bigger fans etc etc. ive no idea how many i killed by chipping the corners or cracking the whole friggin die. i usually bought 4 or 5 at a time to see if i could break some ocing record with them( dont think i ever did). it got expensive but i had alot of fun

 

[freeagent]

462 was pretty tough. I didn't have those rubber pads, and it was still fairly easy to mount up, though I did rush a few times and did end up cracking the corner once or twice.. The last time cost me 100mhz or so. I then later hosed my NF7 and moved to 939. I haven't delidded any of my current CPUs, my coolers clamp down pretty tight.. Actually, I'm curious now.. How thick is the ihs? .005" or so I bet it would be ok. And if it wasn't, that my reason to upgrade

I would imagine it would be like the old days, but instead of a clip you have four screws. Hold the cooler down straight and just tighten up nice and equal. Could always measure from the top of the die down to the pcb and cut a shim with room for compression? I would think as long as the cooler is kept flat against the die it would be ok. My coolers came with a support spacer for lga1151, I wonder if that could be helpful for someone running that socket. I cant afford to replace it right now, so I cant try.

 

[freakdiablo]

I've actually been playing with the idea for the next build whatever it may be - might end up being the next CPU upgrade for my desktop.

 

[Mr.Scott]

Temps don't always decrease when running bare die.
Sometimes you need the IHS to help spread out the heat load. Same principle as more rad. Greater surface area = better temps.
Depends on the platform. Some I get better results bare, some I don't.

 

[Silver Surfer]

I was considering that back with my 3770K but the Intel CPU substrates kept getting thinner and thinner meaning too much pressure can force the center in too much!
You'd be flat lying to say all the CPUs you've run bare die happened problem free, and you know it, and that's one reason why I don't do it.
The main reason I don't run bare die is I run chilled water cooling and I stay on the condensation edge and the IHS actually helps protect the raw die itself from any missed condensation.
Plus probably the last reason is I don't have to.
You can bare die all you want and you still won't be running a 50c P95 26.6 maximum core load temperature on an 8700K with all cores at 5ghz.

- - - Auto-Merged Double Post - - -

I think I run 7 delided cpus neked, I have no issues at all doing this, they all run just fine.
I can't understand taking the risk (very little in my book) of deliding the little lump and just reinstalling the offending IHS, what is the fear?
let me count up what I run buck neked.
4690K
8600K
3, 4790K
3, 8086K
6800K
ok, that comes out to 9 that I run 24/7, buck friggin neked.
why do you reinstall the IHS?

Instead of posting like this, why don't you post how you accomplished these claims with a step by step thread?

Talk is cheap, show proof of what you claim?

When I refer to chilled water cooling my rig is in my sig, with step by step how it was built.

Many on the net that tried running bare die with thinner substrate CPUs had nightmare after nightmare and some even destroyed their CPUs, so what do you know that they did not?

 

[caddi daddi]

oh, I never said I have never had an issue trying to run neked, I have killed my fair share and I am sure I will kill more.

 

[maxfly]

i forgot to metion im going to be going the 8086k route but im too much of a sissy to run it nekkid lol.

 

[caddi daddi]

nothing sissy about it, look at what we have invested in these things, an upgrade from a mid tier rig to a top tier rig is a grand, and that comes out of disposable income not the beef and beans income, and you are going to have a grand or so in a mid tier rig to start from.

 

[Vishera]

I've run 14 soldered de-lidded AMD processors naked. Lots of athlons and semprons from s462 to AM3.

Knowing this is not related directly to de-lidding of Intel processors, from what I do know, AMD is easier to deal with. I was enlightened recently that Intel chips have much less surface area and seem to be fragile if too much pressure is applied. And typically with de-lidded processors, you need a decent amount of clamping force. So the Cpu PCB and or motherboard PCB can bend, which I've seen cause no post issues with my higher TDP AMD processors.

The nice thing about removing layers of copper and thermal paste is you have a much faster transfer of temperature from Cpu core to waterblock. However it seems to have the biggest impact running a much higher TDP than most people would do. I found in most cases an average of about 10c can be lost with running naked. My FX-5000 didn't seem to make much difference as it didn't clock well having the TLB errata issues only reached 2750mhz unlocked. could have done that on an air cooler really.....

I do have a 2700X that some time in the future (no actual deadline set) to de-lid. For now I'm still getting used to the processor and currently running the stock air cooler and heavily limited by temperature near the 1.5v area which I need to surpass the 4.4ghz barrier I'm at now. But then again, it's plenty fast, I run it daily at 4.2ghz typically and leave the voltage on auto believe it or not....

Any Hoot, here's that soldered AMD naked list I've accumulated through the years, just in case some one thinks I've not been down that naked road once or twice. (Have done a couple of Intel chips also )
Here's 14 working cpus and I did Kill 2 so 16 total. Cannot give a direct number on non-soldered processor, we'll say maybe a few dozen or so?

FX-4300
FX-6100
FX-9590
FX-8320
B73 x3
Athon 5000+ unlocker FX-5000
970T
1090T
965BE
940BE
9850BE
9950BE
6400+ x2

I had a 9850! Gosh, I wish I still did so I could throw it under water and crank it up until it screams. If I had a better GPU in that system, I bet it would've soared.