GS65 Repaste Question

Understood, I guess I was referring mostly to the heatsink lottery then.. Wish MSI used something more reliable :/

 

Small update. I repasted it, and did a thin spread this time on the cpu. I also got my undervolt up to -0.165 on the core voltage offset, and my temps are now maxing out at mid 80's playing games with only core 0 and 2 hitting 90 to 91 max during stress tests, with an average of high 60's low 70's. There's only slight thermal throttling on core 2 when the temps peak, and on core 0 and 4 when I stress test it. I think at this point the only things I could do are sand down the heatsink to try to balance it (which I won't since I wouldn't even know where to start on that), or use liquid metal to bring the temps down further. I could technically bump up the fan profile more, but I'm not a fan of jet engines. I might give lm a shot at some point but at the very least, I think it's a large improvement and more than usable at this point. I also managed to fix the creaking on the panels so I'm just gonna jot this whole experience down as a success with room for the improvement in the future.

 

Can you take a clear picture of the underside of the heatsink, if it's not too much trouble?
Thank you. Sanding should be a last resort. I want to see if switching to 0.5mm "Arctic" thermal pads might get cores 0 and 2 down. As this is the same issue many of the 4 core laptops had (due to VRM pad elevation + convex BGA CPU slug + Convex heatsink (you would "think" a very slightly concave (not convex) heatsink would fit a very slightly convex CPU, since that's logic, right? But MSI is smoking heroin or something).

 

Just to make sure, do you mean without the thermal paste? I can do it, I just wanna make sure before I take it out since I already have pictures, just with paste on them. Also it's completely possible that I'm just not good at securing the heatsink properly. This is kind of a first time thing for me, so while I'm pretty stead at applying the paste, I might not be screwing on the heatsink in a way that distributes pressure evenly. If you think better thermal pads might do the trick, I'm down to try that too. This whole thing is a big experiment for me, so as long as I don't fry it(it is my new everyday/work laptop), I'm game for just about anything.

Edit: If they are smoking heroin (or something), I'm mad that they're not including some with every purchase for the rest of us to use lol.

 

You have to repaste if you remove the heatsink anyway, so it doesn't really matter (can't re-use old paste when removing heatsink due to impurities and air bubbles and other stuff. Usually never turns out just right).

What I want to see is the VRM layout of the heatsink. I want to see if the heatsink is also cooling VRM's and if there are thermal pads on the side of the heatsink where the CPU cores are hotter. That would suggest the 'pad' side is elevating the heatsink VERY slightly compared to the "no" pad side.

Try to think of a drawing of two slightly convex surfaces touching each other. Now imagine if one long side gets slightly "elevated up" even by 0.1mm. You can see what happens....

 

I can do that, sure. The pictures I already have don't show the whole design anyway. It'll probably be tomorrow since I'm about to hit a food coma though. Also as far as the vrm's are concered, if they are what I think they are, they are covered by the heatsink, but there's thermal paste on them from the factory, not thermal pads. I've been unsure about whether to repaste those as well or not though, since that's not exactly something I know about, so I've just left it alone in the meantime. I'm wondering now though if that might be part of it as far as the thickness/consistency of the thermal paste vs. the paste that I've been using on the cpu and gpu. IDK, so any input on that before I open it again would be great, as far was whether i should put a dot of paste on the vrm's when I open it again.

 

Guys this is why we have youtube:
Seems like a nightmare to take apart and repaste :x very fragile all over with paper-thin cooling system.... would not want to do that more than once a year.

 

Well now that I see that video, I don't know how much help these will be, but here you are anyway. Also as far as opening it up is concerned, it's really not difficult at all. As long as you don't go hulk on it, it's not bad at all, it just looks scarier than it actually is.

Edit: Not sure why the images aren't being linked in, so just in case, here's an actual link to the images. https://imgur.com/a/PlsS9Ja

 

Very nice.
Tripod heatsink with the single screw on the side with the VRM cooling on it. Less pressure on the top sided cores (possibly cores #0, #2 and #4) than the bottom cores, AND a convex heatsink Well it sure LOOKS Convex !

@Papusan isn't going to be happy to see this.

 

Isn't he the guy who hates all laptops with soldered chips? I don't see why he would like this laptop in the first place

Also small update. When I repasted it this time around, i didn't put any pressure on the heatsink and let the screws do all the work, and it seems to have worked out better than previously. Core differential is down to about 3-4 degrees when gaming, and the highest I saw core 2 hit was 84 degrees. Im sure it'll probably shoot back up to 90ish during stress tests, but at this point, that improvement is probably the best result I'm gonna get unless I lm it, which I'll pass on for now. I need a break from opening up this laptop for a bit lol.