I'd like to thank Dan Rot for alerting me to the fact my post-repaste testing was accidentally done with Integrated graphics. This new screencast runs the same tests with the dGPU enabled: th-cam.com/video/YwbxbyWeoTQ/w-d-xo.html
Hi! I've just repasted my Macbook Pro 15" 2015 following this video. In fact, I found this video back in February or so, but I couldn't get the alcohol until now, and also the Kryonaut delivery took months! Anyway, I followed your original repasting method of spreading, but instead of a thin layer, I put a thicker one, just not so thick that when putting back the heatsink it could spill. The change in performance is really amazing! Thank you so much!
Here are the net results: Cinebench Rendering scores Repaste Paste: Paste: Delta: Normal voltage Old 420cb New 518cb 98cb .75 mv undervolt Old 447cb New 544cb 97cb Undervolt Delta 27cb 26cb 124cb Total increase, undervolt & repaste 124cb = 29.52% 30% Higher score with new paste and undervolt - Same hardware. I am about to do the same drill with the addition of changing the SSD for a 2TB ADATA/XPG SX8200 Pro installed. Will attempt to screen grab the results and video publish. Great Video! Love the style, love the information.
Not sure why you need to manually flatten the paste. Some study (using transparent models) shows you only need to drop the paste in the middle or a line based on the shape of the area, then the pressure between heat sink and the processor will flatten the paste much better than what you do. In fact your way could create bubbles in between causing bad thermal transferring effects.
Please watch from 19:14 and continue watching until my paste job is completed. There you will see what was written in the Thermal Grizzly Kryonaut instructions. It says you can indeed use the spread method (i.e., flatten the paste). But as you can see in my video, Kryonaut doesn't spread easily at all, so all that was left was a faint thin film that was basically nothing at all. In other words, no air bubbles whatsoever can result from a thin film like that. If you cannot see how thin, consider that we are talking less than 0.1mm thin. It's so thin I cannot even measure it. Again, it's basically nothing at all. That is why you see me then change to the X method, which again is described in detail in the manual with photo. Furthermore, transparent model tests have show the X method to be in many ways superior to the dot in terms of getting paste to to the corners.
@@JDW- Yes, Artic also mentions four methods, but finally recommends the One Point method (and one line method because of the heat sink pipe shape). A bubble could be created when two flat plans meet a valley formed by the paste due to manual flattening, and the air has no where to go but to stay in the middle. Of course it is very thin, and is hard to be noticed.
@@太陽雨-i2s As I said in my earlier reply, I used the X method. There is basically no paste on the chip surface from my attempt at the spread method, not even enough paste for microscopic bubbles to form. Again, you can see that clearly in my video. Indeed, what my video shows is that you really shouldn't use the spread method with Kryonaut, even though the spread method is recommended in the official Kryonaut documentation, because the paste just isn't suited for it. Therefore, air and bubbles is a complete non-issue with regard to my video. There are other videos that show methods prone to bubbles, but my video is not one of them.
In reality, it doesn't really matter how you apply the paste. The only important thing is that you put enough on (especially on a bare die). Too much makes a mess, sure, but it wont hurt thermals because the mounting pressure pushes the excess off the edge anyways. This can be an issue if using liquid metal or a conductive paste.
Followed your instructions and renewed the thermal compound on the CPU/GPU of my MacBook Pro mid-2014. It was a very simple job and quick. The advantage I noted after doing this is that the battery has a longer period between charges than before the repasting. I am thinking that the greater heat transfer to the heatsink is not straining the CPU as before and therefore not taxing the battery as much. I downloaded the Intel Power Gadget and observe that the CPU rarely goes above 60C and stays around 50C most of the time. Power consumption is down from 20W or so to 8-10W, even when streaming videos. For me, it was a worthwhile change with huge benefits.
Congratulations, Robert! Thank you very much for making time to share your wonderful story of success. It really is a worthwhile task that’s not too difficult, and I hope more people will try it.
Very nice video. For the thermal paste spreading method, what you want to do is put the paste tube in a Ziploc bag and put the bag into a cup of hot water for a minute or two. The paste becomes much thinner and easier to spread when it's warm.
You mentioned not to wait around for re-pasting until something happens but you leave your batteries this swollen? I am amazed that you are not having trackpad issues with Batteries this one. That’s not just a little bit swollen that’s like something that needs to be taken care of right away and is much more dangerous and risk of leaking those destroying your computer than a little bit of dry thermal paste.
Actually, no. I didn't leave it that way. Within the text description under my video, you will find that I have put a link to "My SWOLLEN BATTERY video." That second video was posted only a matter of days after releasing this repasting video. URL: th-cam.com/video/Om1c-FCc5u0/w-d-xo.html
your video is still being found by people and interested in the advice you said. please tell me the negative impact of this paste is on the crystal. saw people complain about it. they say it leaves marks and scratches the surface. Is it so. what at the moment with the macbook that you are testing on the video?
"crystal"? I'm afraid I don't understand what you mean. I read and reply to every single comment under all my videos, so I also don't know what you mean by people complaining. If you can please be more specific and clear, I can try to answer whatever questions you may have that pertain to my repasting video. Thanks.
@@JDW- my english is bad so you don't fully understand me. I'm asking about thermal paste. I read on forms that it damages cpu and gpu. is it true? How has this paste performed over time? do you recommend to buy? Is there any experience with using another artik mx 4 paste? If so, which one do you think is better. thanks for the answer
Thank you for explaining. I don’t know what forum you’re talking about, but it doesn’t matter. Clearly, those people have no idea what they’re talking about. The paste I used does not damage any CPU or GPU! In fact, Thermal Grizzly Kryonaut is one of the top recommended pastes and has been for many years. I have been using my repasted MacBook Pro to this very day without any problems whatsoever, and the heat and temperature are still under control as you can see in my video. I highly recommend it. And yes, it is better in terms of thermal transfer than MX-4. You don’t need to take my word for that either. Any good, unbiased review of thermal paste will show that.
I noticed you had gfxcardstatus running. During the before test you are using the discrete gpu and in the after test you are using the integrated gpu. This skews the test results because using the discrete gpu increases the temps by about 8 to 10C Celsius. Could you provide test results for the after test with the discrete gpu running?
Thank you, Dan, for noticing that. You're absolutely right. I will run the tests after work tonight and post an update. I really appreciate your keen eye! THANK YOU.
Dan, see my newly pinned Comment for the link to my new screencast. The scores are a bit lower than with Integrated graphics, but such is to be expected since the dGPU will heat the machine more. But both Integrated and dGPU testing shows the scores to be higher than before I repasted. Thanks again.
@@JDW- Great! Thank you for sharing. It helped me to decide to do the repaste too. I hope it'll improve the fan noise while I have an external display connected.
52W is lower than what it is normally and I chose that number because I saw others using it too. Going lower is not a problem but such will reduce your clock speeds accordingly. There is a point below which there is no meaning to cap the wattage because your clock speeds fall lower than the base clock. You could even opt to not cap the wattage at all and just leave Turbo Boost on and undervolt to -75mV. The biggest advantage to using Volta is the undervolt setting. If you really, really want to fine tune all the settings, you would need to set Volta, then test in CineBench, let your machine cool, change the Volta settings, then retest.
How come every tutorial I've read in the past 20 years has said to AVOID ever using a vacuum around electronics because the static build up can cause severe damage, but you suggest using one? @JDW
Perhaps a better question is, in all of those years, how many pieces of electronics have you seen destroyed with your own eyes because a vacuum was used to suck dust out if it? I’ve been working with electronics since the 1980s and I honestly haven’t seen a single circuit board destroyed by a static charged vacuum. Next, if you read through the text description I created underneath my video, and if you watch my video, you will see that I have a rocket blower and a paintbrush prominently displayed for the purpose of cleaning out dust. So my brief mention of a vacuum is rather insignificant seeing that I am not promoting a vacuum as the lone means of cleaning.
I have the same spec MBP and I get 672cb CPU and 63.16fps OpenGL in Cinebench R15 running MacOS Catalina with Volta at -75mV. My cooling paste is still original but I recently opened my Macbook to clean it from dust and found that my batteries are very swollen, a lot worse than yours. But i still get excellent battery life.
Thank you for sharing your experience. First off, it's hard to imagine MacOS Catalina alone giving you such a high CPU score in Cinebench R15. What was your CPU score when running High Sierra (even without Volta)? Next, call Apple now and get your battery replaced. Mine too showed great battery health and run fine off the battery, but if you leave a battery in its swollen state, it technically could burst through your aluminum housing at any time. And whether or not that breakout out of the case will damage the battery, leading to a fire, is also a possibility that needs to be considered. Check your serial number on Apple's website to see if you qualify for a free replacement. More information about battery replacement in my other video here: th-cam.com/video/Om1c-FCc5u0/w-d-xo.html
Do you think that repasting is too big a risk on AppleCare? I bought a refurbished 2015 MBP less than a year ago and opted for the 3-year AppleCare. That said, when I'm running a 4K monitor and watching any kind of TH-cam video or multimedia, my fans are almost always going at a medium-to-high speed, with the accompanying whine.
My AppleCare was expired when I repasted, so your situation is different. Any work you do inside your case is therefore done at your own risk. With that said, so long as you don't damage anything during your repaste and so long as you don't leave any sign you were ever inside the case (stripped screws are a tell-tale sign, as are easy-to-spot fingerprints or paste you may have gotten on an easy-to-see component), Apple would never know you repasted the CPU and GPU and I would think you could safely send the machine in for warranty service under AppleCare in the repasted condition. Reason being, they really cannot know how the stock thermal paste looks versus your repaste. Also, when they replaced my battery, they didn't even remove the heatsink and instead removed the entire motherboard with the heatsink secured. If you are worried, simply don't repaste until your AppleCare is expired. With all of that said, repasting won't keep your CPU from hitting 99.9% because Apple ramps up the clock speed as far as it will go. Repasting just allows you to have higher sustained clocks for longer. In other words, your fans will still blast like mad even after the repaste when your CPU gets hot enough. Apple's software controls that.
If your AppleCare is already over I think there is nothing to lose, assuming you are not going to damage your hardware. If you always load >1080p TH-cam videos that would explain your CPU usage
MBP's are not good with external screens, your fans ramp up quicker to keep the GPU cool. If you were to do the same tasks without external screen (even if the disctreet GPU is enabled) fans run a different profile. I'm talking from experience here.
@@musk3402 Thank you for asking. The answer is "no" because the Gallium in liquid metal thermal pastes reacts with Copper. More specifically, the Gallium is slowly absorbed into the Copper until the Gallium is gone. Therefore, any thermal benefit you'd get over a good traditional thermal paste like Kryonaut would be very short lived. Another reason is because liquid metal must be masked off and therefore is more difficult to properly apply, so it won't accidentally short out surrounding components, versus a traditional non-conductive thermal paste. More more information: bit.ly/32CFMUA
@@musk3402 No because the top of the processor is not copper. Personally, I wouldn't be concerned about liquid metal because Kryonaut is really good paste.
Hey mate I went into apple for a battery replacement and they replaced the entire top chassis for me ass well which was severely bowed, I have always looked after my mac and I just acted as if it was a bizarre occurrence to have my macbook deformed like that out of nowhere. I got it back, cleaned the fans and repasted it so it should be good to go again!
I replaced the thermal paste and when using Safari I have an average of 43°C. On this video it's not even 40°C! What CPU temperature do you people currently have when using Safari typically? Maybe the OS fault?
Thank you for your comment. You didn't mention the type of paste you used. As mentioned in my video and in the text description below the video, I am using Thermal Grizzly Kryonaut, which ranks as one of the best pastes available. Safari temperatures vary wildly based on the content you are viewing. If you watch TH-cam, your temperatures will go up. 43°C is not especially hot for these Macs. They often go well past 70°C a lot of the time, even with good paste, depending on what you are doing. But suffice it to say, if your temperatures are largely unchanged after a repaste, I would be curious to know if you used a cheap paste or a major brand name paste.
hey man thanks for the video i have same exact macbook have since march 2016 do you think i should do a repaste? and with liquid metal or regular paste? i have only done a repaste once on my xbox 360 years ago. thanks!
I used the non electrically conductive Thermal Grizzly Kryonaut because it is safer to use than the Liquid Metal version and because it is pretty much the best non-metal paste you can buy. Any overspill won't post a problem at all, which is not true of liquid metal. Another big problem with liquid metal is that it reacts negatively with Copper heatsinks, and as you can see in my video, copper is used in the MBP heatsink.
@@thegr8rambino Arctic MX-4 is "good enough." It is not as good as Kryonaut, but that's because Kryonaut is the best you can buy, second only to electrically conductive (and rather dangeous) pasts like liquid metal. The temperature difference between MX-4 and Kryonaut isn't huge, and that's why I say MX-4 is good enough. If that's what you bought, you won't have any problem at all. Kryonaut will probably keep your GPU only a couple degrees cooler than MX-4.
I did not use R20 deliberately because at the time R15 was in still wide use so the vast majority of test data out there was done in R15, thereby making R15 testing more useful. And although I could do R20 now, you could also use the older version of R15 on your machine so that you can easily compare with the data I presented.
That advice is sound for other pastes, but it is not applicable for Thermal Grizzly Kryonaut -- the paste I used. I give you all the details about that in the text description under the video, in the section titled "OFFICIAL THERMAL GRIZZLY ADVICE."
great video man! unfortunately there are little to none of this kind, namely undervolting. I owned a 15 inch 2015 base model and I can tell it worked. Highest I went was -95 mV, but i was getting random freezes while idle or low load. -75 mV was stable, so i did not win the silicone lottery. How low did you manage to get without getting any crashes? I am considering getting the exact same model and I am a bit skeptical when i read about the temps. If you do any light gaming under MacOS (DOTA 2 or CS GO), what do the temperatures look like? Thanks :) PS: you just earned a subscriber
Thank you for your kindness in subscribing. I certainly appreciate it. I too tried undervolting lower than -75mV and found instability problems. I leave it at -75mV most of the time, although FCPX prefers I leave undervolting off altogether. FCPX crashes rather often for me, so it's hard to tell when it's undervolting related and just FCPX crashing on its own. But for all other apps, -75mV is working well. But the exact value will vary by machine, due to process differences between the chips. Light or hard gaming doesn't make much difference when it comes to undervolting because you are using Volta to undervolt the CPU only. The GPU is NOT undervolted. I know this for fact because I wrote an email to Gary Matthews who created Volta and he told me that he didn't think it made much difference so he left out GPU-undervolting in Volta. I encouraged him to put the code back in, but that's been a couple months and no reply back from him on that. All said, when you do even light gaming on my high end version of the mid-2015 15" MBP, the dGPU kicks on and the temperature rises. That is because the GPU and CPU share the same heatsink which is tied to the same pair of fans. And when the dGPU is on, even if it's not working hard, you're going to see base temperatures of about 65°C, even on a cool day where the room temperature may be only 20°C. The fact is that these 2015 machines run hot. Repasting is just a start. To keep them fairly cool means to either use room A/C and/or a floor fan aimed at the machine while you're using it. I actually do that, and it drops the machine temps quite a bit.
JDW I saw there is another undervolting method which requires a bit more tinkering. There is a script voltageshift which lets you specify the offset for core, cache, igpu, power limit. Although this takes a lot more testing, I put my money on it that it would be more efficient. If you have any games installed, could you please let me know about temperatures? I am considering buying one again but I don’t remember the temps. Thanks ;)
@@Cristian-em5wc With Volta undervolting at -75mV (and of course with my repasted CPU & GPU), I get up to about 65°C when playing LEGO MARVEL Super Heroes.
Apakah karena panas berlebih, Saat saya melakukan proses editing,di timeline saya terjadi drop frame? Atau lag? Saya menggunakan adobe premiere pro. Tolong jawab.. Terimakasih
Adobe Premier Pro berjalan sangat lambat di Mac. Faktanya, ini adalah perangkat lunak pengeditan video paling lambat yang tersedia di Mac. Itu sebabnya saya menggunakan Final Cut Pro Apple sebagai gantinya.
People might want to make note of why he's using 100% (or at least 90%) iso, you can see how quickly it evaporates in the video. Ya 70% will work but you better wait a bit before connecting the battery if you do use 70% as any standing liquid may cause a short once you add power to the board. Keep that in mind when you take the back cover off to spray with canned air, if you shake it and liquid comes out as often happens, it could short something, trust me, I forgot to unplug the battery when I took the back off once, sprayed inside, some liquid came out when I shook the can while spraying, even though it was powered off, POP 😳 🤦🏻♂
It's important to note I had a squeezable rubber blower blub, not a can of compressed air. With compressed air, frozen white stuff can come out, which is why my video recommends the blower bulb instead. And when I applied 100% ISO alcohol, I use a swab to limit the amount and better ensure it will evaporate quickly. The area must be completely dried before applying new paste. 70% rubbing alcohol has a lot of water in it, so I don't recommend that. It may leave a slight residue too, unlike the 100% or 99% ISO I recommend.
Keep in mind my video was released 4 years ago, and I believe some of the pastes you mentioned didn't exist at that time. I am aware, however, that MX-4 has an 8-year durability, which I find very important. So in that regard, it may be better than Kryonaut, as I believe Kryonaut is only rated for a 3-year life. In fact, I don't believe MX-5 is rated for 8-years like MX-4. What about MX-6? Would you know? Some people don't care about that, but I have grown to appreciate pastes that you can "set and forget" (for a long while).
Kryonaut is designed for overclockers in mind and not for long term use. With that said Mx-6 has the same life as Mx-4 . However MX-6 as 20% better thermals then mx-4.
@@Jimmycozad1980 Thank you for confirming MX-6 has an 8-year life like MX-4, yet with superior performance. That makes MX-6 a winner in my book, even though I have no experience using it yet.
I provide the following link in the text description under my video: th-cam.com/video/Om1c-FCc5u0/w-d-xo.html There's a ton of useful information in the text descriptions I write, so be sure to alway click SHOW MORE to expand them and check out what's there.
Indeed. For that reason, in the text description underneath the video above, I put a link to me other video which talks in depth on that subject: th-cam.com/video/Om1c-FCc5u0/w-d-xo.html
Absolutely. As I mentioned in my follow up video below, I did indeed verify on Apple’s own website that my serial number does NOT qualify for Apple’s battery recall, which is of course a crying shame. My follow-up video is about my experience getting the battery replaced for a fee, so you may want to watch it: th-cam.com/video/Om1c-FCc5u0/w-d-xo.html
Actually, the amount I used was just right. No need to take my word for it either. Check out some TH-cam videos on the repasting pros and you'll see that failing on the side of too much is almost always better than too little. Why? Because the excess will overflow and get out of the way when the heatsink is pressed down firmly. But the bottom line is this. That repasted is still going strong and keeping the temperatures cool. So anyone following my video should have the same success that I did, even if they use their own favorite brand of thermal paste. Speaking of which, MX-4 is an all around good paste that I can recommend. It has an 8-year life versus 3 years for Thermal Grizzly Kryonaut. Kyronaut gets slightly lower temperatures, but for people who don't like repasting ever few years MX-4 is worth a look, and it seems to be better in some regards than MX-5. Happy repasting!
I'd like to thank Dan Rot for alerting me to the fact my post-repaste testing was accidentally done with Integrated graphics. This new screencast runs the same tests with the dGPU enabled: th-cam.com/video/YwbxbyWeoTQ/w-d-xo.html
Hi! I've just repasted my Macbook Pro 15" 2015 following this video. In fact, I found this video back in February or so, but I couldn't get the alcohol until now, and also the Kryonaut delivery took months! Anyway, I followed your original repasting method of spreading, but instead of a thin layer, I put a thicker one, just not so thick that when putting back the heatsink it could spill. The change in performance is really amazing! Thank you so much!
Great to hear about the performance boost, Alex. Thanks for making time to let me know your success!
This video is extraordinarily put together, absolutely amazing! Kudos for another spectacular video.
Thank you for watching and for your kind words, Alex!
Idk it moves really slowwww
Here are the net results:
Cinebench Rendering scores Repaste
Paste: Paste: Delta:
Normal voltage Old 420cb New 518cb 98cb
.75 mv undervolt Old 447cb New 544cb 97cb
Undervolt Delta 27cb 26cb 124cb
Total increase, undervolt & repaste 124cb = 29.52%
30% Higher score with new paste and undervolt - Same hardware.
I am about to do the same drill with the addition of changing the SSD for a 2TB ADATA/XPG SX8200 Pro installed.
Will attempt to screen grab the results and video publish.
Great Video! Love the style, love the information.
Thank you for posting details of repaste, Larry!
Huge thanks for such a thorough and helpful DIY. Exactly what I needed to add thermal paste and pads. Appreciate it!
Thanks so much for posting this ... especially leaving in the section where manually spreading the paste didn’t go as planned.
Not sure why you need to manually flatten the paste. Some study (using transparent models) shows you only need to drop the paste in the middle or a line based on the shape of the area, then the pressure between heat sink and the processor will flatten the paste much better than what you do. In fact your way could create bubbles in between causing bad thermal transferring effects.
Please watch from 19:14 and continue watching until my paste job is completed. There you will see what was written in the Thermal Grizzly Kryonaut instructions. It says you can indeed use the spread method (i.e., flatten the paste). But as you can see in my video, Kryonaut doesn't spread easily at all, so all that was left was a faint thin film that was basically nothing at all. In other words, no air bubbles whatsoever can result from a thin film like that. If you cannot see how thin, consider that we are talking less than 0.1mm thin. It's so thin I cannot even measure it. Again, it's basically nothing at all. That is why you see me then change to the X method, which again is described in detail in the manual with photo. Furthermore, transparent model tests have show the X method to be in many ways superior to the dot in terms of getting paste to to the corners.
@@JDW- Yes, Artic also mentions four methods, but finally recommends the One Point method (and one line method because of the heat sink pipe shape). A bubble could be created when two flat plans meet a valley formed by the paste due to manual flattening, and the air has no where to go but to stay in the middle. Of course it is very thin, and is hard to be noticed.
@@太陽雨-i2s As I said in my earlier reply, I used the X method. There is basically no paste on the chip surface from my attempt at the spread method, not even enough paste for microscopic bubbles to form. Again, you can see that clearly in my video. Indeed, what my video shows is that you really shouldn't use the spread method with Kryonaut, even though the spread method is recommended in the official Kryonaut documentation, because the paste just isn't suited for it. Therefore, air and bubbles is a complete non-issue with regard to my video. There are other videos that show methods prone to bubbles, but my video is not one of them.
In reality, it doesn't really matter how you apply the paste. The only important thing is that you put enough on (especially on a bare die). Too much makes a mess, sure, but it wont hurt thermals because the mounting pressure pushes the excess off the edge anyways. This can be an issue if using liquid metal or a conductive paste.
Followed your instructions and renewed the thermal compound on the CPU/GPU of my MacBook Pro mid-2014. It was a very simple job and quick. The advantage I noted after doing this is that the battery has a longer period between charges than before the repasting. I am thinking that the greater heat transfer to the heatsink is not straining the CPU as before and therefore not taxing the battery as much. I downloaded the Intel Power Gadget and observe that the CPU rarely goes above 60C and stays around 50C most of the time. Power consumption is down from 20W or so to 8-10W, even when streaming videos. For me, it was a worthwhile change with huge benefits.
Congratulations, Robert! Thank you very much for making time to share your wonderful story of success. It really is a worthwhile task that’s not too difficult, and I hope more people will try it.
Very nice video. For the thermal paste spreading method, what you want to do is put the paste tube in a Ziploc bag and put the bag into a cup of hot water for a minute or two. The paste becomes much thinner and easier to spread when it's warm.
Thank you for your kind remarks and for the excellent tip! (So far so good with the "X" application method I chose.)
I've heard it said that spreading is not a good idea as it introduces air bubbles which mess with the conductivity.
You mentioned not to wait around for re-pasting until something happens but you leave your batteries this swollen? I am amazed that you are not having trackpad issues with Batteries this one. That’s not just a little bit swollen that’s like something that needs to be taken care of right away and is much more dangerous and risk of leaking those destroying your computer than a little bit of dry thermal paste.
Actually, no. I didn't leave it that way. Within the text description under my video, you will find that I have put a link to "My SWOLLEN BATTERY video." That second video was posted only a matter of days after releasing this repasting video. URL: th-cam.com/video/Om1c-FCc5u0/w-d-xo.html
your video is still being found by people and interested in the advice you said. please tell me the negative impact of this paste is on the crystal. saw people complain about it. they say it leaves marks and scratches the surface. Is it so. what at the moment with the macbook that you are testing on the video?
"crystal"? I'm afraid I don't understand what you mean. I read and reply to every single comment under all my videos, so I also don't know what you mean by people complaining. If you can please be more specific and clear, I can try to answer whatever questions you may have that pertain to my repasting video. Thanks.
@@JDW- my english is bad so you don't fully understand me. I'm asking about thermal paste. I read on forms that it damages cpu and gpu. is it true? How has this paste performed over time? do you recommend to buy? Is there any experience with using another artik mx 4 paste? If so, which one do you think is better. thanks for the answer
Thank you for explaining. I don’t know what forum you’re talking about, but it doesn’t matter. Clearly, those people have no idea what they’re talking about. The paste I used does not damage any CPU or GPU! In fact, Thermal Grizzly Kryonaut is one of the top recommended pastes and has been for many years. I have been using my repasted MacBook Pro to this very day without any problems whatsoever, and the heat and temperature are still under control as you can see in my video. I highly recommend it. And yes, it is better in terms of thermal transfer than MX-4. You don’t need to take my word for that either. Any good, unbiased review of thermal paste will show that.
I noticed you had gfxcardstatus running. During the before test you are using the discrete gpu and in the after test you are using the integrated gpu. This skews the test results because using the discrete gpu increases the temps by about 8 to 10C Celsius. Could you provide test results for the after test with the discrete gpu running?
Thank you, Dan, for noticing that. You're absolutely right. I will run the tests after work tonight and post an update. I really appreciate your keen eye! THANK YOU.
Dan, see my newly pinned Comment for the link to my new screencast. The scores are a bit lower than with Integrated graphics, but such is to be expected since the dGPU will heat the machine more. But both Integrated and dGPU testing shows the scores to be higher than before I repasted. Thanks again.
@@JDW- Great! Thank you for sharing. It helped me to decide to do the repaste too. I hope it'll improve the fan noise while I have an external display connected.
Hey nice video but would increasing the wattage to 52 would increase the thermal isn't better to lower it ?
52W is lower than what it is normally and I chose that number because I saw others using it too. Going lower is not a problem but such will reduce your clock speeds accordingly. There is a point below which there is no meaning to cap the wattage because your clock speeds fall lower than the base clock. You could even opt to not cap the wattage at all and just leave Turbo Boost on and undervolt to -75mV. The biggest advantage to using Volta is the undervolt setting. If you really, really want to fine tune all the settings, you would need to set Volta, then test in CineBench, let your machine cool, change the Volta settings, then retest.
I love in-depth videos.
How come every tutorial I've read in the past 20 years has said to AVOID ever using a vacuum around electronics because the static build up can cause severe damage, but you suggest using one? @JDW
Perhaps a better question is, in all of those years, how many pieces of electronics have you seen destroyed with your own eyes because a vacuum was used to suck dust out if it? I’ve been working with electronics since the 1980s and I honestly haven’t seen a single circuit board destroyed by a static charged vacuum.
Next, if you read through the text description I created underneath my video, and if you watch my video, you will see that I have a rocket blower and a paintbrush prominently displayed for the purpose of cleaning out dust. So my brief mention of a vacuum is rather insignificant seeing that I am not promoting a vacuum as the lone means of cleaning.
I have the same spec MBP and I get 672cb CPU and 63.16fps OpenGL in Cinebench R15 running MacOS Catalina with Volta at -75mV. My cooling paste is still original but I recently opened my Macbook to clean it from dust and found that my batteries are very swollen, a lot worse than yours. But i still get excellent battery life.
Thank you for sharing your experience. First off, it's hard to imagine MacOS Catalina alone giving you such a high CPU score in Cinebench R15. What was your CPU score when running High Sierra (even without Volta)? Next, call Apple now and get your battery replaced. Mine too showed great battery health and run fine off the battery, but if you leave a battery in its swollen state, it technically could burst through your aluminum housing at any time. And whether or not that breakout out of the case will damage the battery, leading to a fire, is also a possibility that needs to be considered. Check your serial number on Apple's website to see if you qualify for a free replacement. More information about battery replacement in my other video here: th-cam.com/video/Om1c-FCc5u0/w-d-xo.html
Do you think that repasting is too big a risk on AppleCare? I bought a refurbished 2015 MBP less than a year ago and opted for the 3-year AppleCare. That said, when I'm running a 4K monitor and watching any kind of TH-cam video or multimedia, my fans are almost always going at a medium-to-high speed, with the accompanying whine.
My AppleCare was expired when I repasted, so your situation is different. Any work you do inside your case is therefore done at your own risk. With that said, so long as you don't damage anything during your repaste and so long as you don't leave any sign you were ever inside the case (stripped screws are a tell-tale sign, as are easy-to-spot fingerprints or paste you may have gotten on an easy-to-see component), Apple would never know you repasted the CPU and GPU and I would think you could safely send the machine in for warranty service under AppleCare in the repasted condition. Reason being, they really cannot know how the stock thermal paste looks versus your repaste. Also, when they replaced my battery, they didn't even remove the heatsink and instead removed the entire motherboard with the heatsink secured.
If you are worried, simply don't repaste until your AppleCare is expired.
With all of that said, repasting won't keep your CPU from hitting 99.9% because Apple ramps up the clock speed as far as it will go. Repasting just allows you to have higher sustained clocks for longer. In other words, your fans will still blast like mad even after the repaste when your CPU gets hot enough. Apple's software controls that.
If your AppleCare is already over I think there is nothing to lose, assuming you are not going to damage your hardware. If you always load >1080p TH-cam videos that would explain your CPU usage
MBP's are not good with external screens, your fans ramp up quicker to keep the GPU cool. If you were to do the same tasks without external screen (even if the disctreet GPU is enabled) fans run a different profile. I'm talking from experience here.
Fans run in a different profile only if you’re not using a fan control app.
@@JDW- "only if you’re not using a fan control app." i.e. for majority of punters out there.
what a coincidence, i live in Japan too and just ordered the thermal grizzly myself. Great vid!
Why don't you use the coolers to the maximum?
If you mean "why don't I use the FANs set to full speed," the reason is because it would be annoying loud.
Amazing quality, keep up the great work dude.
Thank you!
JDW Do you expect to do a video with liquid metal soon?
@@musk3402 Thank you for asking. The answer is "no" because the Gallium in liquid metal thermal pastes reacts with Copper. More specifically, the Gallium is slowly absorbed into the Copper until the Gallium is gone. Therefore, any thermal benefit you'd get over a good traditional thermal paste like Kryonaut would be very short lived. Another reason is because liquid metal must be masked off and therefore is more difficult to properly apply, so it won't accidentally short out surrounding components, versus a traditional non-conductive thermal paste. More more information: bit.ly/32CFMUA
JDW interesting, so could that possibly ruin the coating of the processor too?
@@musk3402 No because the top of the processor is not copper. Personally, I wouldn't be concerned about liquid metal because Kryonaut is really good paste.
Thats a fantastic piece of kit there! :)
Hey mate I went into apple for a battery replacement and they replaced the entire top chassis for me ass well which was severely bowed, I have always looked after my mac and I just acted as if it was a bizarre occurrence to have my macbook deformed like that out of nowhere. I got it back, cleaned the fans and repasted it so it should be good to go again!
Thank you for sharing your experience! If you don't mind my asking, how much did Apple charge you in total, and what was your turn-around time?
That was probably due to a swollen battery, nothing to do with the other components fortunately.
I replaced the thermal paste and when using Safari I have an average of 43°C. On this video it's not even 40°C!
What CPU temperature do you people currently have when using Safari typically? Maybe the OS fault?
Thank you for your comment. You didn't mention the type of paste you used. As mentioned in my video and in the text description below the video, I am using Thermal Grizzly Kryonaut, which ranks as one of the best pastes available. Safari temperatures vary wildly based on the content you are viewing. If you watch TH-cam, your temperatures will go up. 43°C is not especially hot for these Macs. They often go well past 70°C a lot of the time, even with good paste, depending on what you are doing. But suffice it to say, if your temperatures are largely unchanged after a repaste, I would be curious to know if you used a cheap paste or a major brand name paste.
Hey, it seemed the battery bulged up, maybe problematic.
th-cam.com/video/Om1c-FCc5u0/w-d-xo.html
hey man thanks for the video i have same exact macbook have since march 2016 do you think i should do a repaste? and with liquid metal or regular paste? i have only done a repaste once on my xbox 360 years ago. thanks!
I used the non electrically conductive Thermal Grizzly Kryonaut because it is safer to use than the Liquid Metal version and because it is pretty much the best non-metal paste you can buy. Any overspill won't post a problem at all, which is not true of liquid metal. Another big problem with liquid metal is that it reacts negatively with Copper heatsinks, and as you can see in my video, copper is used in the MBP heatsink.
@@JDW- true and that's really annoying how you have to find out what metal the heat sink is made of thank you sir!
@@JDW- how is arctic mx-4? Is as good as kryonaut? I ordered some with a kit with tools included from amazon but I also want to try the kryonaut too
@@JDW- i thought it only reacts negatively with aluminum heatsinks?
@@thegr8rambino Arctic MX-4 is "good enough." It is not as good as Kryonaut, but that's because Kryonaut is the best you can buy, second only to electrically conductive (and rather dangeous) pasts like liquid metal. The temperature difference between MX-4 and Kryonaut isn't huge, and that's why I say MX-4 is good enough. If that's what you bought, you won't have any problem at all. Kryonaut will probably keep your GPU only a couple degrees cooler than MX-4.
i have got with the MacBook Pro 2012 (i7 3615qm) a score of 552 without changing thermal paste, why is your MacBook so slow ?
is that because the power Limit ?
What version of MacOS and what version of your benchmarking software are you using?
JDW I using Catalina and cinebench r15
It must be the OS then. I am still using High Sierra.
It's a shame you didn't run Cinebench R20, maybe you could do a test using that?
I did not use R20 deliberately because at the time R15 was in still wide use so the vast majority of test data out there was done in R15, thereby making R15 testing more useful. And although I could do R20 now, you could also use the older version of R15 on your machine so that you can easily compare with the data I presented.
I read somewhere that the thermal paste should be the width of a cigarette paper, so the spreading method is best
That advice is sound for other pastes, but it is not applicable for Thermal Grizzly Kryonaut -- the paste I used. I give you all the details about that in the text description under the video, in the section titled "OFFICIAL THERMAL GRIZZLY ADVICE."
great video man! unfortunately there are little to none of this kind, namely undervolting.
I owned a 15 inch 2015 base model and I can tell it worked. Highest I went was -95 mV, but i was getting random freezes while idle or low load. -75 mV was stable, so i did not win the silicone lottery.
How low did you manage to get without getting any crashes? I am considering getting the exact same model and I am a bit skeptical when i read about the temps.
If you do any light gaming under MacOS (DOTA 2 or CS GO), what do the temperatures look like? Thanks :)
PS: you just earned a subscriber
Thank you for your kindness in subscribing. I certainly appreciate it. I too tried undervolting lower than -75mV and found instability problems. I leave it at -75mV most of the time, although FCPX prefers I leave undervolting off altogether. FCPX crashes rather often for me, so it's hard to tell when it's undervolting related and just FCPX crashing on its own. But for all other apps, -75mV is working well. But the exact value will vary by machine, due to process differences between the chips.
Light or hard gaming doesn't make much difference when it comes to undervolting because you are using Volta to undervolt the CPU only. The GPU is NOT undervolted. I know this for fact because I wrote an email to Gary Matthews who created Volta and he told me that he didn't think it made much difference so he left out GPU-undervolting in Volta. I encouraged him to put the code back in, but that's been a couple months and no reply back from him on that.
All said, when you do even light gaming on my high end version of the mid-2015 15" MBP, the dGPU kicks on and the temperature rises. That is because the GPU and CPU share the same heatsink which is tied to the same pair of fans. And when the dGPU is on, even if it's not working hard, you're going to see base temperatures of about 65°C, even on a cool day where the room temperature may be only 20°C. The fact is that these 2015 machines run hot. Repasting is just a start. To keep them fairly cool means to either use room A/C and/or a floor fan aimed at the machine while you're using it. I actually do that, and it drops the machine temps quite a bit.
JDW I saw there is another undervolting method which requires a bit more tinkering. There is a script voltageshift which lets you specify the offset for core, cache, igpu, power limit. Although this takes a lot more testing, I put my money on it that it would be more efficient.
If you have any games installed, could you please let me know about temperatures? I am considering buying one again but I don’t remember the temps. Thanks ;)
@@Cristian-em5wc With Volta undervolting at -75mV (and of course with my repasted CPU & GPU), I get up to about 65°C when playing LEGO MARVEL Super Heroes.
what's your fan control app ??
iStat: bjango.com/mac/istatmenus/
Apakah karena panas berlebih,
Saat saya melakukan proses editing,di timeline saya terjadi drop frame? Atau lag?
Saya menggunakan adobe premiere pro.
Tolong jawab..
Terimakasih
Adobe Premier Pro berjalan sangat lambat di Mac. Faktanya, ini adalah perangkat lunak pengeditan video paling lambat yang tersedia di Mac. Itu sebabnya saya menggunakan Final Cut Pro Apple sebagai gantinya.
@@JDW- maksud saya, apakah karena panas yang berlebihan.
Jadi membuat drop frame, atau frezee?
@@KakakBetaTV Panas yang berlebihan dapat menyebabkan perangkat lunak pengedit video menjatuhkan bingkai.
@@JDW- olraits.. Thanks ma broo..
🙏🏼🙏🏼🙏🏼🙏🏼
People might want to make note of why he's using 100% (or at least 90%) iso, you can see how quickly it evaporates in the video. Ya 70% will work but you better wait a bit before connecting the battery if you do use 70% as any standing liquid may cause a short once you add power to the board.
Keep that in mind when you take the back cover off to spray with canned air, if you shake it and liquid comes out as often happens, it could short something, trust me, I forgot to unplug the battery when I took the back off once, sprayed inside, some liquid came out when I shook the can while spraying, even though it was powered off, POP 😳 🤦🏻♂
It's important to note I had a squeezable rubber blower blub, not a can of compressed air. With compressed air, frozen white stuff can come out, which is why my video recommends the blower bulb instead. And when I applied 100% ISO alcohol, I use a swab to limit the amount and better ensure it will evaporate quickly. The area must be completely dried before applying new paste. 70% rubbing alcohol has a lot of water in it, so I don't recommend that. It may leave a slight residue too, unlike the 100% or 99% ISO I recommend.
Thanks for this video, I really appreciate that.
Amazing video, youre a lifesaver
Apple may replace your battery they did with mine. In aug of 2019 for free. There is a recall due to the swelling issue.
I actually have a video about that here: th-cam.com/video/Om1c-FCc5u0/w-d-xo.html
I use mx-6 better then mx-4 or MX-5.
Keep in mind my video was released 4 years ago, and I believe some of the pastes you mentioned didn't exist at that time. I am aware, however, that MX-4 has an 8-year durability, which I find very important. So in that regard, it may be better than Kryonaut, as I believe Kryonaut is only rated for a 3-year life. In fact, I don't believe MX-5 is rated for 8-years like MX-4. What about MX-6? Would you know? Some people don't care about that, but I have grown to appreciate pastes that you can "set and forget" (for a long while).
Kryonaut is designed for overclockers in mind and not for long term use. With that said Mx-6 has the same life as Mx-4 . However MX-6 as 20% better thermals then mx-4.
@@Jimmycozad1980 Thank you for confirming MX-6 has an 8-year life like MX-4, yet with superior performance. That makes MX-6 a winner in my book, even though I have no experience using it yet.
please get your battery replaced bro, it looks kinda dangerous
I provide the following link in the text description under my video: th-cam.com/video/Om1c-FCc5u0/w-d-xo.html
There's a ton of useful information in the text descriptions I write, so be sure to alway click SHOW MORE to expand them and check out what's there.
Your battery is bulging!
Indeed. For that reason, in the text description underneath the video above, I put a link to me other video which talks in depth on that subject:
th-cam.com/video/Om1c-FCc5u0/w-d-xo.html
👍👍👍
Your battery is swollen apple will replace for free call them before it blows up.
Absolutely. As I mentioned in my follow up video below, I did indeed verify on Apple’s own website that my serial number does NOT qualify for Apple’s battery recall, which is of course a crying shame. My follow-up video is about my experience getting the battery replaced for a fee, so you may want to watch it:
th-cam.com/video/Om1c-FCc5u0/w-d-xo.html
Good to see that you partially screw down the CPU/GPU heat sinks to ensure an even clamp down. Good work! th-cam.com/video/_XQNGNAwrLw/w-d-xo.html
Please use thermal paste sparingly
Actually, the amount I used was just right. No need to take my word for it either. Check out some TH-cam videos on the repasting pros and you'll see that failing on the side of too much is almost always better than too little. Why? Because the excess will overflow and get out of the way when the heatsink is pressed down firmly. But the bottom line is this. That repasted is still going strong and keeping the temperatures cool. So anyone following my video should have the same success that I did, even if they use their own favorite brand of thermal paste. Speaking of which, MX-4 is an all around good paste that I can recommend. It has an 8-year life versus 3 years for Thermal Grizzly Kryonaut. Kyronaut gets slightly lower temperatures, but for people who don't like repasting ever few years MX-4 is worth a look, and it seems to be better in some regards than MX-5. Happy repasting!