I’ve experienced that liquid metal starts seeping into the die which makes it disintegrate. I actually had a CPU that I was cooling with LM blow up after about one year with the solution on it. There was a whole piece of the die that looked like it could have been one of the cores that was just completely gone when I took off the cooler to inspect why it wasn’t booting!
Newer framework machines have switched to something called PTM7958 which is a phase change material that has almost the same cooling performance as liquid metal but it is not conductive which makes it safer in terms of potential short circuits. It also keeps getting better with continued use as the thermal cycling between liquid and gel form (changes at around 55c if i remember correctly) makes it fill microscopic imperfections with use. From what I’ve read this material will last somewhere between 5-10 years before it starts to deteriorate.
I’ve experienced that liquid metal starts seeping into the die which makes it disintegrate. I actually had a CPU that I was cooling with LM blow up after about one year with the solution on it. There was a whole piece of the die that looked like it could have been one of the cores that was just completely gone when I took off the cooler to inspect why it wasn’t booting!
Newer framework machines have switched to something called PTM7958 which is a phase change material that has almost the same cooling performance as liquid metal but it is not conductive which makes it safer in terms of potential short circuits. It also keeps getting better with continued use as the thermal cycling between liquid and gel form (changes at around 55c if i remember correctly) makes it fill microscopic imperfections with use. From what I’ve read this material will last somewhere between 5-10 years before it starts to deteriorate.
It can be worth looking into.