The Slim 5 with LM!

I got around to sorting out the Slim 5 today and it's pretty good, I'll show a few pics and give a brief explanation on the process before getting to the gains.

Here we have the laptop separated from the cooler, first laptop in many I have seen with the GPu and CPU in opposite places too. Putty is on the VRAM and it was actually pretty good, really soft. Pads however didn't look to be making contact or if so, with little to no pressure on the VRMS and mosfets. This is bad if no contact it made as there are no temps you can monitor for these and if these degrade, voltage quality goes to pot and so does your performance.

The red stuff is Thermal Grizzly Shield, a coloured non conformal coating, designed to protect the little capacitors around the dies. LM was firmly rubbed into each cold plate for 10mins each side. I normally bake the cooler at about 95c for a few hours but I thought I'd give the firm rub method a try this time. If you look at the GPU, I have made my own foam gasket to act as a barrier for the LM, the black marks on the cooler are from me putting ink on the barrier and dry mounting it to make sure that it actually makes contact and is effective in it's purpose.

A quick shot of the thermal putty and pot. After this shot I ran a bead of MX-6 around the metal shroud of the 8845HS, this acts an easy none interference barrier since that shroud is high and there is very little room on the substrate to make a barrier of any other kind. The thermal pads were removed from the cooler and everything was mounted up nice.

Now the part you're here for. The thermal and performance results.
Full transparency, this laptop both before and after was tested on my own homemade cooler with 3 Phanteks T30s set to 2000rpm. All runs are with my custom fan curve which is lightly more aggressive than stock but my no means extreme. I chose to test without using a fix fan speed because that isn't really how people would use it in the real world and heck if I used max fans the results would be very much in general favor of LM. The results aren't perfect as I am no lab and I don't have 14 saves in specific areas for a direct apples to apples comparison. Well in the games anyway. The screenshots are massive because I took them in game to show I was being as legit as I could be.

First up Cinebench R23, before then after:

This is an astounding result. +1000 on the score, -10c on the temps, from 85w to near 100w and +200mhz on effective core clock.

Next is a very clear example of apples to apples as I purposely did the most demanding things in DIablo 4, high level Nightmare dungeons, roaming a helltide and a tier 5 Infernal horde. Again, before, then after.

I do believe I reset the HWinfo panel once in the game on these, so even the averages show improvements not just on the peak temps.

Next up is Borderlands 3, The before part of the test was in a lighter area in terms of combat and business, but still demanding enough. I just picked up on the story where I left off in the after, if just so happened to be a little more action orientated.

Still a nice improvement in my humble opinion, especially on the CPU.

Next is Dying Light 2. I love this game and got carried away both times when playing, though again, the story part on the after was more intense. On the before I was just roaming about booting zombies 😄

Last but not least is Dragons Dogma 2. This is an odd one as I had it act up a few times in general, so much so I forgot to reset HWinfo on the before. I think maybe the settings overwhelmed the config, or the driver I'm using was causing issues but I had it crash a few times on the before, but then worked perfectly fine in a more challenging environment in the after. Thing is with Dogma, some areas just thrash the system, but then you can walk into an Inn and the temps dropped like a rock on the after. Averages are not comparable on this one.

So there we have it, over all the CPU REALLY gains from the LM. The GPU and VRAM make some gains, and it does benefit from it. Though I don't think the GPU contact is great on this heatsink. When cleaning it up there was a lot of PTM on the bottom of the die. CPU contact is perfect though.
The delta on the GPU being 10c is likely caused by poor contact of the cold plate. Still, lower temps are lower temps!

I could have made things more conclusive by mirroring the steps in the games other than Diablo 4, D4 was really easy to replicate the load, not so much for the other games. If I had used a more aggressive fan curve on both runs I dare say the gap would have been larger too, or if I used the T30's at their full speeds. This was done in a more realistic manner and I did the testing in a way that I know my girlfriend will use this machine with the fan speeds how she will have them on the laptop and cooler.

Any questions, feel free to fire away 😀