Put new thermal paste on my 1165G7

[Jason_Dagless]

@Allenx51

Few thoughts, maybe try a higher end paste like Thermal grizzly,

I tried that and made next to no difference over the stock stuff. I would say to anyone thinking of trying paste swap to just not bother. The stock paste is good.

I do wonder however, that maybe newer Intel mobile chips just vary up and down so much in heat so quickly that some pastes maybe just can’t keep up with the constant boosts to really high temps and do in fact degrade far quicker than they would with say desktop chips.

Who knows. I get a lot of laptops in and all the Intel 9th gen onwards stuff just sounds like a hairdryer all the time. But if folks are easily led enough to only want super slim machines…

[Allenx51]

@Jason_Dagless

I tried that and made next to no difference over the stock stuff.

I’ll have to disagree. From a temperature standpoint, yeah perhaps, though with these newer intel chips, I’m wondering if the addition of any additional thermal headroom just results in the chips clocking up higher. I just made the swap myself, The thermal pads, and the paste, and saw a decent bump in performance in games, making some of them much more playable.

Gamework (community benchmarks)

Same settings Min:60.35 Max:90.54 Avg: 69.54 Civ IV, Same settings Graphics: 50 FPS AI : 9.00s/turn 2256x1504 low, 4x AA: 41.33 avg across 3 runs (highest was 42.13) 1920x1080 low, 4x AA : 56.67 avg across 3 runs (Highest was 57.04) same settings Winterhold: 37-45 FPS Whiterun : 38-48 FPS Suffice to say, I believe this is a worthwhile bump to performance for the $20-$30 and maybe 30 minutes of time.

But to your point, the stock paste is a lot better than the standard mass production stuff that ships with most laptops. I just had some lying around, and wanted to give it a shot.

[cowpod]

Any recommendations for thermal pastes? Got the pump-out effect using Arctic MX-5 (and MX-4).
Maybe a more viscous one?

[Jason_Dagless]

@cowpod

Got the pump-out effect using Arctic MX-5

Remember less is more with thermal paste. It’s only meant to fill in the tiny microscopic imperfections between die and heatsink. So actually seeing little paste (almost a haze of paste) between the two actually means really good contact and the minimum of paste. You don’t want like a sandwich filling layer of paste between the two.

The reason I think a lot of folks get a little disappointed with their paste change results is that the standard stuff is as good as anything else you can put in without spending silly money. Also sometimes you just get lucky with how you put it back with just the right tension etc. etc.

[Usernames]

@cowpod

Any recommendations for thermal pastes?

Linked below is a post from a user who has learnt of a thermal putty/pad which has been claimed to be comparable to liquid metal with non of the downsides. In a few weeks they will receive them to try out.

i7-1185G7 Reached 60W boost. Happy

Thanks @Usernames and @Second_Coming, I did some more testing and my i7-1165G7 only boosts up to ~38W before thermal throttling which was slightly better than stock (edit: IIRC, also now I’m starting to record temps at an ambient temp of 73F), though it’s possible that the replacement mainboard I received (that I’ve been using for several months) may have been prone to the thermal paste pump-out issue. My Arctic Silver 5 tube is years old, though comparing with other pastes, I’m pretty sure ther…

I had never heard of Shin-Etsu before but looking them up they seem to be as good or better than the well know competition.

[Jason_Dagless]

I’ll be honest here and state a few days ago I actually switched off boosting on my 1165G7 so it maxes at the 2.8GHz.

For what I use it for, there is zero noticeable performance difference but a big difference in fan noise and ramping up and down.

[Usernames]

@Jason_Dagless

For what I use it for, there is zero noticeable performance difference but a big difference in fan noise and ramping up and down.

Thanks for the heads up. I was going to ask what your 28W temperatures were like then I realised without fan speed info it has little value…

I think if the thermal system is able to deal with 60W it should make for a quieter machine with less fan ram overall.

[Jason_Dagless]

Well locked to 2.8GHz average core temp was 70c, max watts was 18.47w and max CPU speed was 2813Mhz.

I did put boost back on and compared fan noise, and I would say a 50% reduction in noise.

[Jason_Dagless]

Had a suspicion something wasn’t right. So I decided today to take the HSF off again and do another re-paste.

Initial pre benches were average max core temp with max boost - 93 degrees. Without boost and max 2.8GHz - 69 degrees.

I took the HSF off and the paste/core/heatsink spread didn’t look right to me. I also took the time to gently use some tweezers to position the thermal pads perfectly (they were a little askew when I got it new). I then applied another dose of Thermal Grizzly and set everything back down and tightened it all up.

Ran the bench again. With Boost max average was 85 degrees (8 degree drop) and at 2.8GHz it 68.75 degrees (I really didn’t expect much difference there).

So I’m pleased with that and it goes back to what I said that sometimes it does come down to the luck of the draw of how it all goes back together as I was using the same paste.

Oh yes and ambient was 27 degrees so not a cool room.

[DannyT]

I decided to move my home setup, I lost my coffee bar by doing so (not much space in my flat) but it was worth it to have a new standing desk and a little more desk space. Plus, bonus, I can use both monitors this way. Still got some cable management to deal with but what do you think? Also swapped the 6GB 1060 GPU for a 12GB 3060.