2020 Air Heatsink Modification Thread

[DanSilov]

did you fully tighten the screw? i am worried that it might pressure the heatsink/motherboard even for the 1.5mm thickness

I’ve fully tightened the screws, yes. Thermal pad is very soft, it doesn’t create that kind of pressure.

 

[kazune_karin]

Hello friends, in the end you will see me around here because I have sold the mac for € 400 damaged and I have given Hauwei a chance to buy something cheap and with a Ryzen that is cooler than Intel ... hahahaha it is an option since in I always have Windows and Mac at home, I still have a Macbook 12 "the Iphones, the Apple wacth, 2 ipad ... in other words, I have not yet left Apple only to pay two laptops now ...
I have had a great time and I hope that my insistence and daring have served some purpose.
See you soon!!!!

Hackintosh is waiting for you?

 

[DanSilov]

I have the same model as you and my GeekBench 5 scores consistently come out at ~1200 single and 3350 multi. Best is 1250 and 3519. That is with no modifications. How come your pre-mod multi-core is so low?

Basically what @Bones13 said. Various reports show that Apple has sorted their assembly issues and latest batches of MBA have much better performance out of the box, to the point where just a thermal pad makes them perform close to the max.

 

[fcracer]

He mentioned he ran them 3 times back to back and took the average. Are you doing the same?

Thank you for that info, I missed that he took a three result average. I’ve re-run the benchmark with the following process so that people can use it as a baseline for their comparisons:

I rebooted my MBA 2020 i7/16/1TB, plugged in the AC power (battery at 39% charge remaining) and ran GB5 five times in a row. Starting temperature for the first run as per Intel Power Gadget (IPG) was 49.33 degrees; IPG was kept running during the benchmarks.

This is for a stock computer:

	Single Core	Multi Core	Time
Run 1	1244	3072	2:40
Run 2	1066	2927	2:48
Run 3	1089	2914	2:48
Run 4	1163	3118	2:40
Run 5	1117	3108	2:42
Average	1135	3028	2:44

As a control, I let the computer’s fans come back down to idle speed and the CPU temp down to 49 degrees (took a fews minutes) and ran the benchmark one more time:

	Single Core	Multi Core	Time
Run 1	1227	3052	2:41

What I find interesting about this benchmark is how variable it is. In comparison, Cinebench R20 gives almost the same results each time. A few days ago, I could get 3300 consistently from GeekBench, but today, even the first pass from a cool CPU was only 3072.

The ambient temperature today is higher which may be a factor and the battery is charging which may also be a factor. I’ll have to run the same five passes again when the battery charges to 100% to see if it makes any difference.

 

[DanSilov]

What I find interesting about this benchmark is how variable it is. In comparison, Cinebench R20 gives almost the same results each time. A few days ago, I could get 3300 consistently from GeekBench, but today, even the first pass from a cool CPU was only 3072.

The ambient temperature today is higher which may be a factor and the battery is charging which may also be a factor. I’ll have to run the same five passes again when the battery charges to 100% to see if it makes any difference.

These are much better results than what I had with my stock MBA'20 i7/16, but the inconsistency is the clear sign of the thermal limitations — as soon as the chassis/heatsink gets hot, CPU is throttled. With the shim mod this is remedied to a degree, and the overall performance becomes more stable, longer and more sustained, but eventually you still hit the ceiling. But this only applies to certain tasks, like gaming of video rendering, and for regular tasks you will be okay 95% of the time.
[automerge]1592291474[/automerge]

This is for a stock computer:

	Single Core	Multi Core	Time
Average	1135	3028	2:44

So, with both mods (shim+pad) you can achieve +15% performance for single core and +30% performance for multi-core.

 

[Loog]

So, I've finally got to repasting my MBA'20 i7/16 with Kryonaut and replacing my original shim with the one from @Loog.

Loog has sent me two versions, 0.5mm and 0.3mm, so I've tested both to see which performed better. Before the installation I've polished both shims to practically mirror-like state just for the sake of removing all the potential imperfections out of the equation.

View attachment 924372

If anyone wants to know, I've used P2000, P2500, P3000, P5000 and P7000 grit sandpaper one after another. I've also polished the heatsink, but just a little bit.

Kryonaut thermal paste is a bitch to work with due to viscosity, compared to Arctic Silver. It takes longer to apply thin even layer on top of the CPU die, and it gets everywhere around the CPU, but since everyone says it's the best, I had to try.

Testing process remained the same, clean macOS, nothing running except for the Intel Power Gadget and Geekbench 5.1.1. Waiting for the CPU to cool down to ~40°C and then run the test. I've run each test 3 times, so below is the average of 3 runs.

Now, the numbers. Here's the comparison table:

	Pre-mod	Original shim, Arctic Silver, thermal pad	New shim (0.5mm), Kryonaut paste, thermal pad	New shim (0.3mm), Kryonaut paste, thermal pad
Geekbench (Single-core)	1169	1265	1261	1269
Geekbench (Multi-core)	2621	3849	3804	3939
Cinebench R20	903	1139	1144	1169

As you can see, 0.5mm shim performs slightly worse than 0.3mm, but all in all the difference in negligible. I do like that the top multi-core result is the whopping 50% increase compared to the pre-mod!

Thermally the new 0.3mm shim+Kryonaut paste performs better, which can be seen on the following Cinebench R20 temperature curves comparison:

View attachment 924415

In fact, during the whole Cinebench R20 run the temperature never reached 94°C, and stayed below 90°C for the most of it (with the thermal pad of course). Compared to the pre-modded version running at constant 100°C I'd say it's a huge improvement.

One thing that bothered me is that I couldn't get my Geekbench multi-core score above 4000, while even the i5 models have recently managed to climb over it. I've run all the previous tests on macOS 10.15.4, so I didn't update to the 10.15.5 for the resting to make sure it didn't affect the outcome. But after all the tests were done, I've updated the macOS and ran Geekbench 3 again a few times. Below is the average of 4 runs.

	macOS 10.15.4	macOS 10.15.5
Geekbench (Single-core)	1269	1300
Geekbench (Multi-core)	3939	3994

But one of those runs gave me the following result:

View attachment 924405

Finally! And it looks like the single-core performance in macOS 10.15.5 has slightly improved as well.

Great review @DanSilov and assumption proved that 0.3mm is better for the later models and 0.5mm for the first generation Retina units. Also an interesting observation that the custom shim is slightly more efficient than the 15mm x 15mm 0.3mm 1.5 shim split equivalent, I assume down to lightly more contact area with the heatsink, albeit marginal. Thanks for taking the time to test both and provide excellent feedback.

 

[fcracer]

What I find interesting about this benchmark is how variable it is. In comparison, Cinebench R20 gives almost the same results each time. A few days ago, I could get 3300 consistently from GeekBench, but today, even the first pass from a cool CPU was only 3072.

The ambient temperature today is higher which may be a factor and the battery is charging which may also be a factor. I’ll have to run the same five passes again when the battery charges to 100% to see if it makes any difference.

For the sake of completeness, I conducted the tests again with 100% battery and got the same results (~1135 single core and ~3055 multi). However, I did manage to replicate the very high multi-core score from last week by waking the computer from sleep and immediately running Geekbench. In that situation, the i7/16/1TB produced this result:

	Single Core	Multi Core	Time
Run 1	1227	3447	2:32

Therefore, I would suggest for consistency when doing before/after comparison, get the laptop hot by running Geekbench at least one time, and then take the five subsequent runs and average them. That should produce a consistent before/after for comparison.

The Arctic pad arrives on Friday so I'll do another set of before/after and report the results here. I'm hoping to get 3,800 multi-core using the same test method as the above, which would represent a 24% increase in sustained performance; this would bring it mighty close to my 2018 13" MBP i7 which gets ~4,100.

 

[kinchee87]

Great review @DanSilov and assumption proved that 0.3mm is better for the later models and 0.5mm for the first generation Retina units. Also an interesting observation that the custom shim is slightly more efficient than the 15mm x 15mm 0.3mm 1.5 shim split equivalent, I assume down to lightly more contact area with the heatsink, albeit marginal. Thanks for taking the time to test both and provide excellent feedback.

The "split shim" was also pasted with Arctic Silver, which might not be as good as the Kryonaut paste. I can't find the thermal conductivity of the Arctic Silver tho ...

 

[srkirt]

Hello friends!! just a note ... just taken out of the box and installed W10 and downloaded Cinebench 20.
Surprised for the performance of AMD Ryzen 4-core, Removing the screen that is Full HD and looks sad next to the MBA the € 629 laptop is surprising !!! as Apple does not wake up the Chinese eat it ...

 

[dipdop123]

For the sake of completeness, I conducted the tests again with 100% battery and got the same results (~1135 single core and ~3055 multi). However, I did manage to replicate the very high multi-core score from last week by waking the computer from sleep and immediately running Geekbench. In that situation, the i7/16/1TB produced this result:

	Single Core	Multi Core	Time
Run 1	1227	3447	2:32

Therefore, I would suggest for consistency when doing before/after comparison, get the laptop hot by running Geekbench at least one time, and then take the five subsequent runs and average them. That should produce a consistent before/after for comparison.

The Arctic pad arrives on Friday so I'll do another set of before/after and report the results here. I'm hoping to get 3,800 multi-core using the same test method as the above, which would represent a 24% increase in sustained performance; this would bring it mighty close to my 2018 13" MBP i7 which gets ~4,100.

Does the chassis get noticeably hot(er) after running Geekbench a few times?
[automerge]1592347376[/automerge]

Hello friends!! just a note ... just taken out of the box and installed W10 and downloaded Cinebench 20.
Surprised for the performance of AMD Ryzen 4-core, Removing the screen that is Full HD and looks sad next to the MBA the € 629 laptop is surprising !!! as Apple does not wake up the Chinese eat it ...

It is more of a Intel issue I think. That's why Apple is probably announcing the transition to ARM on Monday

 

[christophosphorus]

This has been a fascinating thread and I've yet to complete reading it, BUT:

Thank you all for what you do, and what you contribute!

I have a 2020 i7 MBA, and with the work-at-home policies as of late, I'm finding that any trivial amount of CPU usage makes the fans turn on and eventually throttle, especially when I have an external monitor. I've resorted to disabling Turbo Boost and limiting the fan speed because it's near impossible to use video conferencing (VC) software without the laptop sounding like a flying saucer about to take off.

I'm in the process of gathering the required mod list and will report back with my results. Hopefully I can contribute a small fraction to what everyone here has already done.

Cheers!

 

[Loog]

This has been a fascinating thread and I've yet to complete reading it, BUT:

Thank you all for what you do, and what you contribute!

I have a 2020 i7 MBA, and with the work-at-home policies as of late, I'm finding that any trivial amount of CPU usage makes the fans turn on and eventually throttle, especially when I have an external monitor. I've resorted to disabling Turbo Boost and limiting the fan speed because it's near impossible to use video conferencing (VC) software without the laptop sounding like a flying saucer about to take off.

I'm in the process of gathering the required mod list and will report back with my results. Hopefully I can contribute a small fraction to what everyone here has already done.

Cheers!

@christophosphorus, this has been the reason a number of us have undertaken these mods to help exactly with the issues you are seeings. I suspect you are a Zoom, MS Teams, Webex user, all which play havoc with these devices over prolonged periods of time. Some to greater or less degrees.

One thing I would suggest is you try removing your docking station and see if this makes any difference as part of your trails, I find my device runs a little warmer with this when in use, also when on mains too. Also be cognisant when TimeMachine runs, if set to auto over wireless all of these factors add certain loads that other more casual users will not see or have any impact on their workflow if not using video conference, also consider not using virtual backgrounds which also have negative effects and ensure that GPU is enabled on services like MS teams.

The mods do help in all these areas, however some other subtle changes to how you work may also help out a little here as well.

 

[christophosphorus]

@christophosphorus, this has been the reason a number of us have undertaken these mods to help exactly with the issues you are seeings. I suspect you are a Zoom, MS Teams, Webex user, all which play havoc with these devices over prolonged periods of time. Some to greater or less degrees.

One thing I would suggest is you try removing your docking station and see if this makes any difference as part of your trails, I find my device runs a little warmer with this when in use, also when on mains too. Also be cognisant when TimeMachine runs, if set to auto over wireless all of these factors add certain loads that other more casual users will not see or have any impact on their workflow if not using video conference, also consider not using virtual backgrounds which also have negative effects and ensure that GPU is enabled on services like MS teams.

The mods do help in all these areas, however some other subtle changes to how you work may also help out a little here as well.

Appreciate the advice, Loog. I've been experimenting with disabling certain features or apps to see what effects they will have on the energy use; it's been decent so far. I will absolutely test without the monitor/dock first, but I'll also run tests with the monitor attached just so I can see real-world changes for my personal usage.

VC can be devastating on these laptops, and I'm sure I won't experience a 100% silent laptop. But, even with disabling Turbo Boost and limiting my fan speed to 3-4,000rpm, I'm avoiding throttling. That's good news to me, and it makes me hopeful that these mods you have all outlined here will help even more.

Above all, I just want my laptop to run normally without peaking at 100*C and throttling over the smallest of tasks. It's such a capable machine that has been handicapped by the thermal design, and that bothers me conceptually more than anything. I have a Windows machine I built that's much more capable to handle intense tasks, but I'm still chasing the MacBook mods and am very glad I came across this thread last night!

On a side note, I'm still waiting for PM approval so I can send you a request for your custom copper shim. I'm not able to send messages yet, but am interested if you still have the 0.3mm available.

 

[PhightinPhils26]

I've been lurking on this thread... though at 76 pages deep, it's more like a rope... since the beginning of May. I have to tell you, it has not been easy to follow. I started working on a Google spreadsheet with my results and was curious if you think anyone else would benefit. A lot of times I see (what I consider) eye popping results, but then realize that it was posted by a user that doesn't have the same specs as me. For someone like myself, where I'm a bit hesitant to throw a shim in my brand new machine, I thought it might be easier to put data in one place.

Here's what I have so far, and it should be read only. Does anyone else see value in something like this, or am I like James Mattingly II and alone with my principles?

 

[srkirt]

In the end I think there will be a change ... MBP 2015 fixed with a new retina display in exchange for the MBA2019.

 

[Loog]

Appreciate the advice, Loog. I've been experimenting with disabling certain features or apps to see what effects they will have on the energy use; it's been decent so far. I will absolutely test without the monitor/dock first, but I'll also run tests with the monitor attached just so I can see real-world changes for my personal usage.

VC can be devastating on these laptops, and I'm sure I won't experience a 100% silent laptop. But, even with disabling Turbo Boost and limiting my fan speed to 3-4,000rpm, I'm avoiding throttling. That's good news to me, and it makes me hopeful that these mods you have all outlined here will help even more.

Above all, I just want my laptop to run normally without peaking at 100*C and throttling over the smallest of tasks. It's such a capable machine that has been handicapped by the thermal design, and that bothers me conceptually more than anything. I have a Windows machine I built that's much more capable to handle intense tasks, but I'm still chasing the MacBook mods and am very glad I came across this thread last night!

On a side note, I'm still waiting for PM approval so I can send you a request for your custom copper shim. I'm not able to send messages yet, but am interested if you still have the 0.3mm available.

@christophosphorus, all 0.3mm shims have shipped, I still have a few 0.5mm, the difference is marginal, @DanSilov did a review on both versions. If you want one of these and you're willing to do a review for the forum drop me a PM when you can and I'll get one out in the post for you FOC.

 

[christophosphorus]

@christophosphorus, all 0.3mm shims have shipped, I still have a few 0.5mm, the difference is marginal, @DanSilov did a review on both versions. If you want one of these and you're willing to do a review for the forum drop me a PM when you can and I'll get one out in the post for you FOC.

I appreciate that. Seeing that the 0.3mm was preferred for the 2020, I just placed an order for a set via Amazon. Would have loved to use yours, though! No matter.

Regardless, I'll be doing a writeup and reporting back here.

 

[kinchee87]

I've been lurking on this thread... though at 76 pages deep, it's more like a rope... since the beginning of May. I have to tell you, it has not been easy to follow. I started working on a Google spreadsheet with my results and was curious if you think anyone else would benefit. A lot of times I see (what I consider) eye popping results, but then realize that it was posted by a user that doesn't have the same specs as me. For someone like myself, where I'm a bit hesitant to throw a shim in my brand new machine, I thought it might be easier to put data in one place.

Here's what I have so far, and it should be read only. Does anyone else see value in something like this, or am I like James Mattingly II and alone with my principles?

This sort of thing would be a good way to document everyone's results and to build some sort of consensus on what mod variant to go for! It's too hard to rely on people adding their mod to their signature, especially if they change their mod over time.

 

[silvia_18]

Hi, just a quick question - what's the best method to remove the black paste Apple has used on the heat sink? Thanks everyone.

 

[PhightinPhils26]

This sort of thing would be a good way to document everyone's results and to build some sort of consensus on what mod variant to go for! It's too hard to rely on people adding their mod to their signature, especially if they change their mod over time.

I don't mind hosting this on my Google Drive. I just added a new column for username. I'm not sure opening this up to the free world with a single edit link is a good idea, but I'm up for suggestions.

 

[IngerMan]

I think there are 2 scenarios. 1 like myself stock MBA is cool 32c on wake up. It runs in the 40’s to 60’s for office work. We know from experience with the Gen 10 it bursts all the time taking the temps up and down,

#2 your starting 20c hotter.

Stock I’m GB5 1300, 3600 for my i7 16 256.

that tells me my Stock thermal paste is good and I don’t need to change it.

Others in the 2900-3200 range tells me the thermal paste is not good For a i7.

You have to open activity monitor before testing and should be 98% plus cpu. If you have 20% Kernal Panics you will have bad scores. It’s common for Kernal panic to peak at times. Also Time Machine will kill a test.

thermal pad only will increase multi score 400 points. Nothing on single core...

 

[Jays1022]

Can someone with stock and modded air tell me how's opening 10-15 tabs on chrome for the computer? I just ordered i5/16 and was wondering if chrome is okay? I need to use chrome's extension to do exams (recording for 2 hours)... I need a mac that wont freeze or break down on me...

 

[BLiT_Racing]

Can someone with stock and modded air tell me how's opening 10-15 tabs on chrome for the computer? I just ordered i5/16 and was wondering if chrome is okay? I need to use chrome's extension to do exams (recording for 2 hours)... I need a mac that wont freeze or break down on me...

I use Chrome almost exclusively. Stock, 12 tabs would run fine, excluding Facebook (their new UI is buggy AF) or YouTube. Modded, I can literally run anything without issue. (Sans Facebook bugs, of course.) I realized the other day that I was accidentally streaming 4K on YouTube and still didn't have any thermal issues.

 

[PhightinPhils26]

If you would like to participate in keeping better track of Mods and Results, here is an Edit URL (Against my better judgement) to a Google Sheet. If you have any questions please let me know. Please don't do us dirty and put false results in.

 

[silvia_18]

If you would like to participate in keeping better track of Mods and Results, here is an Edit URL (Against my better judgement) to a Google Sheet. If you have any questions please let me know. Please don't do us dirty and put false results in.

Maybe we could all agree on the same method/steps for when we do benchmarks? (which ever people think is best)