Intel finally responds to bending 12th gen Alder Lake CPUs on LGA1700 motherboards

kac77

2[H]4U
Joined
Dec 13, 2008
Messages
3,314
I guess Intel has managed to create their own Red Ring of Death a-la Xbox 360...........if you never embraced that console and only know the meme, it's because of the x-clamp retaining system that Microsoft went with that, over time, caused the PCB to warp as the console heated up and then cooled down, you'd get microfractures in the tracings and, well, there you go.

I suspect MB manufacturers will immediately start including back plates.....or, also likely, hello "Rev. B" board designs.....
 
Last edited:
Article states bending is caused by the socket retention mechanism, doesn't matter what cooler you're using. Shady shit by Intel, hello AMD!
And the way to avoid said bending?
Is to use a solid(metal usually) backplate that wont allow the retention mechanism to warp the socket or bend the mb once the cooling solution is in place. So unless you know how to make your own backplate...
 
Has anyone here seen this issue personally? If so post some pics and share what you have done to either fix it or RMA parts and what that experience is like. I’ve been running a 12700k , Tuf D4 board and Noctua U12S Chomax since shortly after launch. My cpu socket retention is Lotes and not Foxconn. The Noctua 1700 kit includes a nice backplate. Recently I did a case swap and am not seeing any warping of my board or socket after 3-4 months. I haven’t removed the Noctua or installed any other cooler since the original build. Just sharing my experiences here to add some data.
 
I commented in another thread on the subject. I had boards with both the Lotes and C/A retention brackets(I don't know what company this is but I haven't seen a Foxconn) I noticed that the Lotes was better constructed and has a slightly higher Z height that puts less pressure on the CPU. I to use the Noctua 1700 kit and coolers without issues.
 
Last edited:
.....or, also likely, hello "Rev. B" board designs.....
Due to the seemingly-widespread issues people are having with dozens to hundreds of WHEA errors a second when using PCIe 4 video cards, those are already coming. It's affecting all the mobo makers and Gigabyte is already working on a new revision. I'm sure the others are too, but I've only been following them, as that's whose board I got. They're going to be RMAing like crazy and are probably going to take a bath on this.
 
And the way to avoid said bending?
Is to use a solid(metal usually) backplate that wont allow the retention mechanism to warp the socket or bend the mb once the cooling solution is in place. So unless you know how to make your own backplate...
The ILM already has a solid metal backplate…

The only way to fix is to use washers to lessen the force on the cpu.
 
  • Like
Reactions: pavel
like this
Has anyone here seen this issue personally? If so post some pics and share what you have done to either fix it or RMA parts and what that experience is like. I’ve been running a 12700k , Tuf D4 board and Noctua U12S Chomax since shortly after launch. My cpu socket retention is Lotes and not Foxconn. The Noctua 1700 kit includes a nice backplate. Recently I did a case swap and am not seeing any warping of my board or socket after 3-4 months. I haven’t removed the Noctua or installed any other cooler since the original build. Just sharing my experiences here to add some data.
No issues here with my MSI board. Been running ADL since the beginning.
 
Has anyone here seen this issue personally? If so post some pics and share what you have done to either fix it or RMA parts and what that experience is like. I’ve been running a 12700k , Tuf D4 board and Noctua U12S Chomax since shortly after launch. My cpu socket retention is Lotes and not Foxconn. The Noctua 1700 kit includes a nice backplate. Recently I did a case swap and am not seeing any warping of my board or socket after 3-4 months. I haven’t removed the Noctua or installed any other cooler since the original build. Just sharing my experiences here to add some data.

If you haven't removed the cooler, I don't think you'd notice it.

Personally, I have the Strix Z690-A D4. I went from a 12900k to a 12600k to another 12900k. Each time I remove my AIO from the socket, I notice that the thermal paste is significantly thicker in the middle of the IHS near the socket retention point showing it is pushing it down more in the middle than on the edges. I don't know if it affects temperatures, but my experience has been that the issue is real in that the middle of the IHS has significant more bend than the sides.
 
So do the spacer mod - which is?: using 1mm, M4, washers? That's all your need? Where can I get those on Amazon? There's hundreds - various nylon & stainless steel....
Edit: Plus, you need a torx screw driver?
 
Why are people making a big stink about this now, when it has literally existed since the beginning of Intel LGA sockets in 2004/2005? I have not had a single Intel board since then that didn't suffer some form of warping or IHS deflection, bar OEM motherboards from Dell and HP where they screwed the heatsink directly into the case frame. Two old examples being the HP DC5800 and the Dell Optiplex 7xx.

Intel's stock cooler with the plastic push pins puts a huge strain on the motherboard because there's no support mechanism. You're literally relying on the thickness of the PCB for support, which was never designed to take such loads.

I've had many a Intel board die from socket warping, so many so that I stopped swapping heatsinks once they were installed. Because when you relieve the pressure is when the cracks in traces and solder joints happens. Once you have that tension applied, you can't let go because everything is stretched at that point and will break when you let it relax.
 
Why are people making a big stink about this now, when it has literally existed since the beginning of Intel LGA sockets in 2004/2005? I have not had a single Intel board since then that didn't suffer some form of warping or IHS deflection, bar OEM motherboards from Dell and HP where they screwed the heatsink directly into the case frame. Two old examples being the HP DC5800 and the Dell Optiplex 7xx.

Intel's stock cooler with the plastic push pins puts a huge strain on the motherboard because there's no support mechanism. You're literally relying on the thickness of the PCB for support, which was never designed to take such loads.

I've had many a Intel board die from socket warping, so many so that I stopped swapping heatsinks once they were installed. Because when you relieve the pressure is when the cracks in traces and solder joints happens. Once you have that tension applied, you can't let go because everything is stretched at that point and will break when you let it relax.
Whoa whoa whoa! You can't state those facts; they completely go against the narrative. ;)
 
Whoa whoa whoa! You can't state those facts; they completely go against the narrative. ;)

Don't worry, Intel will have somewhere to shift the blame when LGA AM5 comes out. "hey hey now, let's talk about AMD and how their AM5 socket is warping CPUs, look at those noobs that have no experience in LGA socket design over our 17 years of industry experience!"

I expect to see this in a leaked internal Intel slideshow sometime in the future.
 
Don't worry, Intel will have somewhere to shift the blame when LGA AM5 comes out. "hey hey now, let's talk about AMD and how their AM5 socket is warping CPUs, look at those noobs that have no experience in LGA socket design over our 17 years of industry experience!"

I expect to see this in a leaked internal Intel slideshow sometime in the future.
After I experienced that Intel plastic push pin heat sink I was gob smacked that very few mentioned just how bad it was.

With regards to the future it's going to be interesting to see if AMD makes the same mistakes.
 
After I experienced that Intel plastic push pin heat sink I was gob smacked that very few mentioned just how bad it was.

I used to buy boxes and boxes of those push pins to replace smashed/broken ones on customer machines. I go through a lot fewer of them now because people know stock Intel coolers are garbage and/or Intel doesn't ship coolers with a good number of their CPUs anymore and people buy non-crap heatsinks.
 
The ILM already has a solid metal backplate…

The only way to fix is to use washers to lessen the force on the cpu.
If you're using a cooler with the 1700 kit - it kind of raises everything? Do you need the washers even if you have a 1700 kit with your cooler?
 
Has anyone here seen this issue personally? If so post some pics and share what you have done to either fix it or RMA parts and what that experience is like. I’ve been running a 12700k , Tuf D4 board and Noctua U12S Chomax since shortly after launch. My cpu socket retention is Lotes and not Foxconn. The Noctua 1700 kit includes a nice backplate. Recently I did a case swap and am not seeing any warping of my board or socket after 3-4 months. I haven’t removed the Noctua or installed any other cooler since the original build. Just sharing my experiences here to add some data.
It's really confusing - the way I understand it - if you watch Bullzoid's video on youtube - a lot of the aftermarket coolers already have a concave (slightly - you might need to try a level to see the gap?) plate which allows to make contact with the cpu. There is less and less FLAT heatsink plates - which would invite the issue - as the retension mechanism the Intel 1700 has introduces too much tension on everything that causes the bend in the first place.
What I want to know and NO ONE seems to mention - is if these 1700 mount brackets these coolers either have included or if you go through the process of buying or receiving in the mail - ease the tension eliminating the problem.
I was going to use the Dark Rock Pro 4 - and I received what I think is the 1700 mount kit (it only consists of two parts) and I guess I use the rest of the Intel hardware (screws etc.) in the Be Quiet cooler box?
I downloaded the manual from the Be Quiet website - and it looks like it raises the mount mechanism a little? Is this equivalent to the washer mod or will the washer mod still be needed?
I think this 1mm increase raises the level enough so the force/tension is just a little less avoiding the bend. Again, the problem seems more related to a flat plate heatsink so when that is used, it's putting tons of pressure on the cpu.

https://www.igorslab.de/en/intel-vs...ader-really-what-is-better-ryzen-9-or-core-9/

https://www.tomshardware.com/news/easy-mod-reduces-alder-lake-cpu-temperatures-5-degrees-celsius

None of these guys/authors talk about using 1700 mounting kits - and if that helps at all.
 
Why are people making a big stink about this now, when it has literally existed since the beginning of Intel LGA sockets in 2004/2005? I have not had a single Intel board since then that didn't suffer some form of warping or IHS deflection, bar OEM motherboards from Dell and HP where they screwed the heatsink directly into the case frame. Two old examples being the HP DC5800 and the Dell Optiplex 7xx.

Intel's stock cooler with the plastic push pins puts a huge strain on the motherboard because there's no support mechanism. You're literally relying on the thickness of the PCB for support, which was never designed to take such loads.

I've had many a Intel board die from socket warping, so many so that I stopped swapping heatsinks once they were installed. Because when you relieve the pressure is when the cracks in traces and solder joints happens. Once you have that tension applied, you can't let go because everything is stretched at that point and will break when you let it relax.
I've never had this issue with Intel's LGA sockets in the past. I've been using them since 2007.
 
I've never had this issue with Intel's LGA sockets in the past. I've been using them since 2007.

Just because you haven't seen/had it happen to you, doesn't mean it never happened.

Here, I brought out my GA-965P-DS3 with an E7200 and unmounted the Intel cooler so you can see.

Warped PCB under the socket:


Warped concave IHS:


As I stated earlier, been a problem since the introduction of LGA sockets.
 
Meh. As long as they keep functioning it really isn't an issue. It would already be out there if this was a real problem.
It really isn't anything that affects performance in any meaningful way, just something for them to work the kinks out of for Gen13 or whatever they use for their new process then.
 
Meh. As long as they keep functioning it really isn't an issue. It would already be out there if this was a real problem.

Warped motherboards have definitely caused issues with LGA sockets in the past. LGA775 boards usually just outright died when a BGA connection failed, but larger sockets like LGA1366 and LGA2011 would have memory channels drop out before eventually having a socket failure.

Ironically, the motherboards with the highest socket failure rates I've encountered over the years are Intel Desktop Boards. I have a few of them in my scrap pile that can be made to briefly work by hot airing the socket with flux, but it doesn't last because the solder Intel used was garbage.
 
I remember people swapping 1366 retention brackets because some were really flakey on good boards lol. Those Intel boards were garbage too like gigabite said, utter complete trash, there's a reason Intel stopped making them lol. Their servers weren't much better though, we had a few Intel E5-V2 servers boards fail multiple times and ended up replacing them with dells or hps.
 
I remember people swapping 1366 retention brackets because some were really flakey on good boards lol. Those Intel boards were garbage too like gigabite said, utter complete trash, there's a reason Intel stopped making them lol. Their servers weren't much better though, we had a few Intel E5-V2 servers boards fail multiple times and ended up replacing them with dells or hps.
I remember it being blamed on the weight of the enormous CPU coolers we were using .... they were like half the size of the things we are having to use now.
 
Warped motherboards have definitely caused issues with LGA sockets in the past. LGA775 boards usually just outright died when a BGA connection failed, but larger sockets like LGA1366 and LGA2011 would have memory channels drop out before eventually having a socket failure.

Ironically, the motherboards with the highest socket failure rates I've encountered over the years are Intel Desktop Boards. I have a few of them in my scrap pile that can be made to briefly work by hot airing the socket with flux, but it doesn't last because the solder Intel used was garbage.
Everybody had shit solder at that point, the sudden removal of Lead and Mercury from the products resulted in metric shit-tons of subpar products.
 
Everybody had shit solder at that point, the sudden removal of Lead and Mercury from the products resulted in metric shit-tons of subpar products.

While everyone had problems with ROHS solders, some were far worse than others. Intel, Pegatron (ASUS, ASRock), Nvidia and ECS/PC CHIPS were on my shit list at the time. While most just used shitty ROHS solder, Pegatron was just miserable. They used traces and VIAs probably 25-50% smaller than everyone else and printed silk screen masking backwards on all components. So not only were components a royal PITA to remove and replace, you constantly had to remind yourself that everything is backwards from virtually every other manufacturer.

Nvidia, Pegatron and ECS boards were so bad that you could sometimes induce BGA failure with your index finger just lightly pressing down on the PCB surrounding whatever ASIC it was. Nvidia would induce their own bond wire failures on the 9xxx series laptop GPUs without any input from the user. I remember this one ASUS OEM board where I could poke the corner of the north bridge and cause the IGP to draw colored jail bars on the screen lol.
 
What kinds of ROHS solders did they use? My current job uses SN96 and sometimes SN95 for ROHS products. We use good old SN63 for most products. I dread it when we use SN18. It is quit difficult to work with and has a bad tendency to crack during thermal shock testing.
 
What kinds of ROHS solders did they use? My current job uses SN96 and sometimes SN95 for ROHS products. We use good old SN63 for most products. I dread it when we use SN18. It is quit difficult to work with and has a bad tendency to crack during thermal shock testing.
At that time ROHS anything was rushed, I don’t remember what they were called, if they were called anything. It was one of those rules that smacked them out of left field so most manufacturers were not prepared for it, many had to rush forward with faulty or incomplete formulas because they just didn’t have the time they needed to test and refine them.
 
What kinds of ROHS solders did they use? My current job uses SN96 and sometimes SN95 for ROHS products. We use good old SN63 for most products. I dread it when we use SN18. It is quit difficult to work with and has a bad tendency to crack during thermal shock testing.

No idea what was used, but I suspect it may have been one or more mixtures that had copper, gold or silver. Those metals are known to work harden from heat and mechanical stress.

I remember a conversation I had with a Keysight engineer back in 2017 about garbage ROHS solder and he was telling me about all of the horrible problems they were having trying to find something that would replace lead that wasn't either brittle or hard or both. To that day, they couldn't. There's really no other metal or amalgam of metals that can replicate exactly the properties of leaded solder.

I'd argue that the banning of lead didn't have the desired effect of reducing hazardous waste, it created far more of it. Instead of the ROHS initiative, they should have done an anti-plastic initiative, which is a far worse problem.
 
No idea what was used, but I suspect it may have been one or more mixtures that had copper, gold or silver. Those metals are known to work harden from heat and mechanical stress.

I remember a conversation I had with a Keysight engineer back in 2017 about garbage ROHS solder and he was telling me about all of the horrible problems they were having trying to find something that would replace lead that wasn't either brittle or hard or both. To that day, they couldn't. There's really no other metal or amalgam of metals that can replicate exactly the properties of leaded solder.

I'd argue that the banning of lead didn't have the desired effect of reducing hazardous waste, it created far more of it. Instead of the ROHS initiative, they should have done an anti-plastic initiative, which is a far worse problem.
I don't think the bending has been a serious problem except on select mobos like some of the X58 ones. And honestly the motherboard can be bent by people overtightening it to the mounts just as easily. I'd say it's rare for me to see a 3-5+ yo board that does not have some bend in it.
 
I don't think the bending has been a serious problem except on select mobos like some of the X58 ones. And honestly the motherboard can be bent by people overtightening it to the mounts just as easily. I'd say it's rare for me to see a 3-5+ yo board that does not have some bend in it.

There is no "adjusting" a stock Intel cooler. It's either not mounted or SEND 'ER ALL THE WAY.

And while there certainly were rigid aftermarket coolers, the vast majority had preload springs either as part of the screw itself, or as a kind of leaf spring on the mount.

No matter if Intel or aftermarket, the pressure was going to be immense.

But my original argument still stands, IHS and motherboard warping are not new phenomenons, they've been around for a very long time.

At least it wasn't like Socket 462 where you had to heave your entire body weight on the retention spring and risk slipping and punching a hole straight through the motherboard, or at bare minimum, ripping chunks and components off of it.
 
There is no "adjusting" a stock Intel cooler. It's either not mounted or SEND 'ER ALL THE WAY.

And while there certainly were rigid aftermarket coolers, the vast majority had preload springs either as part of the screw itself, or as a kind of leaf spring on the mount.

No matter if Intel or aftermarket, the pressure was going to be immense.

But my original argument still stands, IHS and motherboard warping are not new phenomenons, they've been around for a very long time.

At least it wasn't like Socket 462 where you had to heave your entire body weight on the retention spring and risk slipping and punching a hole straight through the motherboard, or at bare minimum, ripping chunks and components off of it.
I will say the newer 115x intel HSF do not put that much pressure on the board. Some other HSF definitely put more pressure down.
 
At least it wasn't like Socket 462 where you had to heave your entire body weight on the retention spring and risk slipping and punching a hole straight through the motherboard, or at bare minimum, ripping chunks and components off of it.
LMAO. OMG this was the scariest thing dealing with that socket. I've never applied so much pressure using a flat head screwdriver in my life.

Once it was on it was on. There was no IHS either so if you got that CPU to boot it was an act of God considering the amount of pressure applied to the CPU package.
 
Back
Top