I found this and thought it was worth sharing:
No need to waste life for YT mumbling.
White silicone.
Period.
Mike
1 Thank
They kind of mention it. But the ones that I have tried that win the tests are also almost impossible to spread evenly. I got some of this Frost X45 paste. I used half of the tube in attempts to get a thin even coating on the CPU. It wanted to stick to whatever I was using to spread it more than the CPU. I finally got it done and it does work well. But I will use something else next time.
As to white silicon, I have used it for years on heatsinks for electronic components (power diodes, regulators, mosfets, etc) it works just great. When you are trying to cool something with a small surface area and high power dissipation, there are better options. Sometimes reducing temps by that 1 or 2 degrees C to keep a processor from going into thermal throttling is worth using something more exotic.
2 Thanks
The best thermal interface materials for flashlights are Phase Change Material (PCM) pads like PTM7950.
1 Thank
I was dealing with devices 20-30 years old, where white silicone was as fresh as at the day it was applied. Try that with modern super-duper thermal paste.
Mike
1 Thank
Its all about what the application calls for. I have never gone for more than 3 to 4 years between CPU swapsā¦ As I mentioned, keeping the processor from frying itself is more important than the paste being good in 20-30 years.
If I wanted 20 to 30 years in a low power situationā¦yup, I may just use white silicon. No doubt, great stuff. I was just looking through all of the different thermal compounds that I have here. Guess whatā¦ I found a big old tube of that white silicon heat sink paste!! 
Maybe it is just not state of the art for CPUs or even flashlights where maxing out cooling potential is the priority.
1 Thank
āWhite siliconeā?
Yāmean the rubbery stuff like good olā Fujik, or stay-gooshy paste like AS5?
Haha, ay-ess-five looks like itās spelling out āassāā¦ 
Reason itās confusing term is that rubbery silicone like RTV is chains of R2SiO, and the gooshy stuff is called āsilicone heatsink compoundā and the like, so itās kinda ambiguous.
1 Thank
Always used liquid metal thermal pastes in my high performance lights, specifically Thermal Grizzly Conductonaut.
DC2Pro seems like another good option now as well.
Objectively better than any regular āpasteā thermal compound but only really necessary if youāre trying to get the last 1% of performance out of something.
That is the one ā¦ at least that is how I understood @sp5it to be referring toā¦ It was definitely what I was talking about. I donāt think silicon RTV stuff would be very good for heat transfer.
1 Thank
Itās the trade-off between good usable life expectancy and good thermal performance. That stuff lasted forever but its thermal performance was ass. The fancy modern stuff performs well but gets brittle soon.
4 Thanks
There is no substantive difference in thermal paste types. Arctic this and arctic that is just selling jargon.
2 Thanks
Truth. For 95% of users and use cases, the differences between the $30 TIM and $5 TIM is negligible. Itās all in what you need it to do. For flashlights, you want good performance and longevity (how many of us open a flashlight to replace the paste?). Iāve had good luck with Arctic MX-4 and Silver 5.
2 Thanks
That is the key. But in general heat is about the biggest enemy of electronics. Many are willing to pay to get the additional 10 to >15% of heat transfer. There are lots of tests that show that kind of difference in various thermal pastes. It can be the difference in a CPU going into thermal throttling or not.
In any case folks using BLF are quite used to doing things to improve lights that the majority of people simply donāt care about. Using the best possible material for heat transfer should be easily understood around here. It is for me. Especially when the difference in the cost for the material is relatively low.
As I said before, Iāve just stopped caring about thermal pastes and just use stuff like PTM7950.
As long as itās not a high-strung light, Fujik would not only hold the stuff in place, but squoosh out the air between, and at least that was better than air gaps from irregular, bumpy, or simply non-flat parts.
Doesnāt take much for a star to just be resting on parts of the rim if theyāre shaped like Pringles. And some pills were notorious for being bowed, or having the nipple in the middle from lousy machining poking up the star and keeping it from making good contact.
So even Fujik was better than nothing.
1 Thank
I had never heard of it. Looking it up it is definitely better at conducting heat than airā¦ but not much. Pretty close to regular RTV silicon sealer.
"Fujik thermal paste typically has a thermal conductivity of around 0.92 W/mĀ·K "
āThe thermal conductivity of air at atmospheric pressure is approximately 0.0262 W/(mĀ·K) at 20 Ā°C.ā
The Frost X45 paste that I got claims :
Thermal Conductivity (W/mĀ·K) 5.1
Some of the liquid metal compound like those mentioned above claim .
Thermal Conductivity (W/mĀ·K) around 15
But they are conductive so you really have to be careful with them or you can get the magic smokeā¦
But the best case is metal to metal contact with only the thinnest layer of paste just sufficient to to fill minor irregularities in the mating surface. That may be why the real crazies always lap mating surfaces. I think that applications that are overly thick may be worse than using nothing.
1 Thank
Yeh, and 50Ć better conductivity over air gaps is a huge increase.
āWay better than nothingā vs āthe bestā are two entirely different things.
Fujik cures and is permanent. It doesnāt dry up, get brittle, or flake off, nor does it have to be reapplied every coupla years.
It is what it is.
True it is conductive, and also will eat anything that is aluminum haha
If you have lapped surfaces you sometimes actually get better thermal transfer with no paste or other material between them, the molecules are close enough together to transfer heat directly between surfaces 
But yes youāre also correct, if the paste is thick or you donāt have enough pressure (or both) performance can decrease significantly regardless of what paste conductivity is.
One advantage of liquid metal is that it is very low viscosity compared to pastes, so you donāt really have to worry too much about that.
1 Thank
Iāve been using PTM7950 pad on my CPU and GPU too, but in a flashlight!? Iām not sure if itās worthy.
1 Thank
If thereās one thing to learn about thermal conductivity ratings, it is that most of them are hot trash.
1 Thank