Well I actually personally used toothpaste as thermal grease. I'd heard about it for years. I recently moved and couldn't find any of my thermal grease after I'd taken my cooler off to clean and replace the thermal grease because temps were getting quite high. It lowered temps about 20*C over the crusted on garbage that was on this old PC I'm using. Worked fine for 3 days while I waited to get some Arctic Silver MX-4 from Amazon (perils of living in a small town). Was surprised how small effect on temps was going from toothpaste to MX-4 as well. Only lowered temps 3-6*C under full load burn testing. Just thought some might find it interesting.
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Real Life Tested Toothpaste as Thermal Grease.
- Thread starter Logan M
- Start date
drescherjm
[H]F Junkie
- Joined
- Nov 19, 2008
- Messages
- 14,941
Interesting. I wonder how long the temps would hold up.
I replaced it after 4 days total. I saw no degredation in cooling performance. I gamed on it extensively and ran the Intel burn test right after applying it and right before I removed it. Not sure how long it holds up though as 4 days is a relatively short amount of time.
Susquehannock
Supreme [H]ardness
- Joined
- Jul 26, 2005
- Messages
- 4,614
Assuming you used the standard white paste. No surprise there. Have done this myself. Contains high amounts of Alumina (aluminum dioxide) which is a key component in many TIMs. Main issue is larger and inconsistent particle size in the toothpaste. Typically works well until the fluids dry up. Then temps are not that desirable. Of course, YMMV.
Does it make the system smell minty fresh?
It actually did still smell minty when I cracked it open and cleaned it off to replace. Kinda nice and definitely not a smell I'd associate with working on PCs.
Spartacus09
[H]ard|Gawd
- Joined
- Apr 21, 2018
- Messages
- 1,930
I vaguely remember someone testing this, as long as you're not putting constant load it'll be ok for a couple months before it dries up.
LOL. Its all about costs. Saving 10 cents over millions of units adds up.
Nobu
[H]F Junkie
- Joined
- Jun 7, 2007
- Messages
- 10,051
Pretty much this. Intel knows people will buy it as long as it works and doesn't fail immediately in stock configuration. If they have to modify the chip to get better performance, that's a minor nuisance compared to the alternative (a slower chip that can't be OC'd to the same performance). Of course, intel may have to rethink that stance soon, but for now I think even with the competition they'll stick with what they've got as far as interface material between die and heatspreader.
I did this back in the day when I didn't have a license or debit card or know of any place to pick up as5. Worked fine. Think it actually ran a little bit once heated. This was on a 2800+ Barton. I don't hesitate to run without paste for testing of hardware. Running a 1700x at stock on an amd cooler I see like 60c cinebenching with no paste.
Yeah if my system wasn't on one of those old shitty Intel push and twist heatspreaders I'd have run without paste too. Those things just barely make contact with the cpu. Total trash design.