Overclock.net banner
Cancel
Create thread New Forums

Thermal Putty vs Thermal Pads: Comparison Chart

Tags
test thermal pad thermal putty tim
4 reading
Jump to Latest
21 - 40 of 186 Posts
#21 ·
Ichirou said:
The only concern now is longevity. No two brands/products are the same, and some might not last as long as others, leading to earlier thermal degradation.

Not too big of an issue if the GPU uses the stock cooler and can be detached every so often to replace the compound, but in a water loop, longevity is paramount.
Click to expand...
For me TG PP10 has lasted for years but if it's compressed enough (for me about 1mm) it dried out and becomes chalky, but temps never changed noticeably.
 
#22 ·
Blameless said:
I had wondered if the Upsiren UX Pro was the same stuff as the EVGA putty...looks like they probably are.
Click to expand...
Ordered some after seeing his chart this morning. Lots of old stuff around here that needs repasted/etc.
 
Save
Reply Quote Rep+ Rep+
#24 ·
Ichirou said:
The only concern now is longevity. No two brands/products are the same, and some might not last as long as others, leading to earlier thermal degradation.

Not too big of an issue if the GPU uses the stock cooler and can be detached every so often to replace the compound, but in a water loop, longevity is paramount.
Click to expand...
I've been cooking some CX-H1300 13.5W on a "hotplate" at 125c for almost 1000 hours now, it's about 1.25-1.5mm thick. It's gotten a bit more stiff/dry when cool, but if I scrape it off and knead it together it's still soft enough to reuse.

Higher temps cause the oil to bleed out faster because of the reduced viscosity. If you're using good putty on a waterblock, I expect it will last a very long time because of the relatively low temps.
 
Discussion starter · #25 ·
Sector-z said:
@snarksdomain can you test them to ? Thanks in advance 🙂

UPSIREN UX Ultra PRO 16.8w/mk Thermal Putty
Fehonda LTP81 18w/Mk Liquid Thermal Putty
Teucer 3D Graphite Thermal Pad
Upsiren 24W/MK Thermal Pad
Zezzio 16.8W/MK THERMAL PADS

View attachment 2655400
Click to expand...
I've tested UX Pro Ultra, and have 3D Graphite pads on hand. I've got both Fehonda putties on the way, as well as the 24w/mk Upsiren pads. Will look at getting the Zezzio pads
 
Discussion starter · #30 · (Edited)
Sector-z said:
Ok sorry I thinked that there haved a "Pro" & "Pro Ultra"

For know what is the best scenario "possible" for the Top chart, usung the best Thermal Putty + Copper shim like 1mm would be great too. It will give us idea what we miss if we don't do it.
Click to expand...
A lot of people get mixed up when talking about UX due to the naming convention.

Below is an Obsolete chart that gives an approximate comparison of the effect of using shims with putty. This chart needs to be taken with a grain of salt, as I did not track or log ambient temperatures when I made it.

The gap between VRAM and cooler is 1.8mm and for the shims I used 1.6mm and then ~0.1mm of putty on either side. TH949-1 is not an ideal putty to pair with 1.6mm shims on this card as it is on the stiffer end of putties and has a hard time squeezing out. I also likely used too much and you can see the core temp is higher due to non-ideal core contact.

I will be testing shims and putty again once I've got more pads on the chart.

Sometime this year I will be retesting everything again with a different videocard (possibly an AMD 6950XT). This is because all the current Nvidia cards only show 2 degree increments for the Memory Junction Temperature, something I realized 4-5 days ago. AMD 6000 series cards show down to 0.1 degree increments, while the 7000 series seems to also only show 2 degree increments. I had hoped to upgrade the test card with a 50 series Nvidia, or the next gen AMD, but may be forced to use AMD 6000 series instead unfortunately.

Font Slope Rectangle Parallel Electronic device
 
#32 ·
Interesting how the increased thermal transfer from the memory to the heatsink causes the core to be higher, that's surprising to me considering how much more mass overall is dealing with the heat, perhaps the shims just aren't thick enough.. only solution is more shims and more putty lol
 
Discussion starter · #33 ·
Dragonsyph said:
What’s the ambient temp for each test? Where they the exact same?
Click to expand...
Ambient temps were different for each test and fluctuate during the test as well for the hour of data that I use for these results. A more detailed chart is available to look at and you can even see the processing of data on some of my streams.
 
Discussion starter · #34 ·
I actually think thinner shims are the solution. I believe the core wasn't getting a good direct contact and the stiffer putty made that harder to achieve. Each VRAM module only puts out somewhere between 3.4 and 4 watts (assuming more power based on OC and seeing temp increase based on OC).

We'll find out when I retest shims. I'll be more careful to not use too much putty ND also have some thinner shims I'll test too.
 
#35 ·
Travis Scott said:
Interesting how the increased thermal transfer from the memory to the heatsink causes the core to be higher, that's surprising to me considering how much more mass overall is dealing with the heat, perhaps the shims just aren't thick enough.. only solution is more shims and more putty lol
Click to expand...
The overriding difference is the GPU contact pressure, not the heat produced by the memory. As snarksdomain says the shims were probably too thick.

At the other end of the spectrum the core temps are lower on the stock pads because performance is being throttled.
 
  • Rep+
Reactions: snarksdomain and StAndrew
Save
Reply Quote Rep+ Rep+
#36 ·
snarksdomain said:
Ambient temps were different for each test and fluctuate during the test as well for the hour of data that I use for these results. A more detailed chart is available to look at and you can even see the processing of data on some of my streams.
Click to expand...
Thanks mate, this data has been valuable. Thank you for the hard work, have a blessed day.
 
  • Rep+
Reactions: snarksdomain
Save
Reply Quote Rep+ Rep+
#37 ·
Welcome aboard Snark 🫡
 
  • Rep+
Reactions: T.Sharp
Save
Reply Quote Rep+ Rep+
#38 ·
Blameless said:
The overriding difference is the GPU contact pressure, not the heat produced by the memory. As snarksdomain says the shims were probably too thick.

At the other end of the spectrum the core temps are lower on the stock pads because performance is being throttled.
Click to expand...
If possible you can weld and machine the copper shims to the existing heatsink, large gauge copper wires can be used too.
 
#39 ·
How are the copper shimms used?
IC|putty|copper|putty|cooler, or how should I imagine that?
 
  • Rep+
Reactions: Travis Scott and snarksdomain
Save
Reply Quote Rep+ Rep+
#40 ·
I would love for Arctic or Thermal Grizzly to launch a putty; that way you know what you’re getting each time.

In the mean time, I’ll pickup some UTP-8.
 
  • Rep+
Reactions: snarksdomain
Save
Reply Quote Rep+ Rep+
21 - 40 of 186 Posts
About this Discussion
185 Replies
34 Participants
Last post:
Melan
Overclock.net
posts
28M
members
562K
Since
2004
The premier forum for overclocking experts and enthusiasts. Discuss hardware optimization, custom builds, benchmarking, cooling solutions, and pushing the boundaries of computing performance. From beginner guides to extreme overclocking, join our technical community to master system tuning.
Full Forum Listing
Explore Our Forums
Intel CPUs Hardware News NVIDIA PC Gaming Water Cooling