I don't normally interact with these videos, but I really appreciate the effort you put into this. Please test the thermalright solution when you can! Also, maybe try cutting the squares into a more accurate size, or ever ever so slightly less than the full coverage as it still spreads out. Should have less overspill and may result in better temps.
Thank you for your comment. That's a good idea on undercutting it a bit. Will definitely try that out. I used to do that with pastes, where I would spread it and then wipe a 1/8" (3mm) border clean of paste and still have great contact patch.
Use a plastic spatula or guitar pick to drag along the edges of the die or IHS where the PTM overhangs, it will cut the PTM and make it super easy to peel the film without lifting or tearing. I still don't think it matters at all if there are some air bubbles, since they should just squeeze out once there's heat and pressure. DerBaur has also stated as much, if you don't believe me.
dude thanks, I was figuring out similar idea U shape but start from one edge. Mine pad should arrive in couple weeks and I try your style first :) I got an MSI p65 laptop. CPU is power limited but I hope I can drop temps and by that drop the fan noise. If it cools my laptop I slap it also on my desktop that probably is due for thermal paste replacement as its been running since ryzen 1700 release without any re-paste :D still has the stock thermal paste pre applied on the wraith cooler so I imagine huge cooling boost :p
I gave you thumbs up, as it's pretty smart application here, BUT...air bubbles will be ironed out by immense heat generated within the cpu, and it will do it darn quickly. Similar like protective glass bubbles on the phone go away after normal usage, and that screen barely gets warm. I would say this is a safe bet to stop being paranoid over some bubbles. We are good. Just the pressure of the cpu mounting/waterblock + heat, and you don't have to do summersaults around the application methods. That's my experience. I've put on protective films on glass as well, and heat does its' work.
Nope, tried everything, did this literally yesterday and it was my second attempt at this. I out it in the freezer for a while to make it easier too. Nope, still teared in several places. It's like it's glued to the foil. Of the 50EUR I paid for it, I maaaaybe have enough left for another application. It was a nightmare and I really am not an graceless elephant when it comes to stuff like that. I had to patch it up, but then it was so uneven and thick in some places, that it was locking my PC when I wanted to run Cinebench for example. I had to take the cooler off, scrape the excess PTM, put it on again and now it's working as intended. But man... it's the worst.
It's cool that more companies are releasing PCM. I know that there is still Laird Tpcm 7125 7,5 W/mk. I wonder if Thermalright Heilos 8,5 W/mk will also be PCM, but so far I have not seen such information. Thanks for the test 👍
I'm hoping to buy some Helios soon, but haven't found an online retailer yet. Will check out that Laird PCM when I can. Edit: In my opinion it likely is, but only time will tell.
@@jamesdavies686 I had compared 3 of those on Testbench 2.1 but decided to stop core/CPU testing on that testbench as I didn't feel it was accurate enough to trust the results, with such a small temperature span. I'll soon be testing all of these things, and more PCM's on Testbench 3. I don't plan to buy the PTM7950 from LTT as it will be the same as the one from Moddiy. I also have some sheets coming from Caplinq. I do have Gelid, Thermal Grizzly and Maxtor PCM's though, among others. Testbench 2.1 will continuue with Putty/Pad/Shim testing for now so we can fill out the chart a bit more (mostly shim+putty combos). All the current charts are located at the Google Drive link below (Test Chart Repository)
Great and decent info man, without the extra misleading stuff other channels do/add, keep up the good work, and keep improving your channel, you have a decent informational channel here that I would trust more than the big channels. Good luck.
Hey there, with PTM thermal pad and my experience I found that sticking it in the refrigerator for around a half hour or so makes application a hell of a lot easier. Idk what your house ambient temp is, but cooling it in the fridge definitely helped me out. It made peeling the plastic applicator extremely easy and the pad remained whole.
This is a solid tip for sure. I've done the fridge technique a couple times with good success. Some Phase Change sheets now have overhanging plastic films on both side which makes first peel super easy. And second one just as easy. For the sheets without the extra plastic film, those little sticker tabs certain help. Can put one on either side of plastic at one corner and peel apart without much fuss. Another cool trick, and this one from T.Sharp, is using a credit card or other firm plastic holding it at a 45 degree angle and running it along the edge of the IHS/Die with modest pressure to "chop" off any excess. This makes second peel go smooth and doesn't lift a flap at the corner. Several others have also suggested undercutting the square by a couple mm on each side as this will use less of a large sheet and give more applications. This won't be a huge benefit for small sheets though (where it's only large enough for 2 CPU applications.
Nope, tried everything, did this literally yesterday and it was my second attempt at this. I out it in the freezer for a while to make it easier too. Nope, still teared in several places. It's like it's glued to the foil. Of the 50EUR I paid for it, I maaaaybe have enough left for another application. It was a nightmare and I really am not an graceless elephant when it comes to stuff like that. I had to patch it up, but then it was so uneven and thick in some places, that it was locking my PC when I wanted to run Cinebench for example. I had to take the cooler off, scrape the excess PTM, put it on again and now it's working as intended. But man... it's the worst.
@@EpicKillstreak is it tearing when you peel the first plastic film? Or is it tearing after you've pushed it down onto the CPU and ran your finger over it, and the peeling second film?
@EpicKillstreak that's no fun. It can be tricky to get it to peel. I had the best luck with the 0.25mm phase change sheets, and T.Sharps method of running a plastic spatula around the edge at a 45 to kind of cut it.
I tested the upsiren pcm 0,02mm vs the ptm7950 from aliexpress (0,25mm), specifically from passionate girl store and this was a little better than the upsiren. The only downside with these pads is when you take them out (specially on laptops because of the small bare dies) after some months I made a rookie mistake and ended up cracking the cpu die on my msi laptop because i tried to remove the heatpipe when the laptop was cold, so it's better to warm up everything if you need to dissasembly your laptop. Confirmed this a couple of days ago, used some pcm on copper sheets, warm'em up and then let them cool down and dude it was hard af to separate them without heat. Anyway great channel, love your videos, a lot of information we don't find anywhere else, thanks!
Phase Change materials definitely become hard once they cool down (any temperature below 45C of the material itself will be solid). If anyone uses Phase Change on an AM3 or AM4 CPU they will need to use the AMD recommended "Twist" method if/when removing the cooler. This goes for dried paste as well.
Before i replace thermal paste with PTM my theory is a D-Limonene cleanup (Arcticlean is an option, step 1 bottle is D-Limonene i think) followed up with 99% alcohol thorough clean for the silicone residue from the old paste increases the adhesion and stability, esp inside the micro-pores from the cold-plate, dabbed on alcohol must lay flat on the surfaces without moving away or forming droplets, that usually means the surface is still hydrophobic from silicone residue. On heatsinks i use a small gas torch (comparable with a torch cigarette lighter) with short bursts, it oozes out the remaining silicone or whatever that liquid is inside the pores, wipe off with a piece of paper shop towel. Done this on 100's of heatsinks. After that the new material fills up those empty voids without silicone interference. On car spraypainting silicones are really nasty stuff, 1 droplet of silicone on a panel (tire shine and such) and the respray is trash.
That's a great tip. I've used a little butane lighter a few times, and it's surprising to see how much residue oozes out. I will look into D-Limonene, thanks for the tip.
A fun experiment to try and would be interesting to see the results. Instead of using the usual thermal pads on Vram/Ram chips, can you experiment with using PTM or PCM instead (even double layer it) and see if it gets good results.
It sounds like a cool idea, but that would be stacking up to achieve 1.8mm gap (so about 8-9 layers of PCM material. If/when it melts, it would then basically glue the VRAM to the Cooler and would be extremely difficult to pull apart. This would also be more expensive than putty or pads. I mostly just use a high end Thermal Putty on VRAM/VRM these days, rather than pads. I might try shims and PCM on a card I'd be willing to sacrifice someday. Just worried about damage when trying to pull it back apart.
@@snarksdomain yeah it would be cool to experiment on an oIder card you don't care about anymore, with either shims or PCM. Definitely very curious to see what happens.
Just found your channel (obviously new to me...lol). Your methodology seems quite thorough, so kudos for that. Looking forward to your future installments! Keep up the great work.
Thank you, and welcome to the channel. Lots of fun stuff to come. I'll be testing out some different viscosities of the Upsiren Putties soon too. Will likely start that testing once I'm back home.
@@snarksdomain Awesome, very interested in your results. Have you done any work with the Thermal Grizzly KryoSheet? Roman (and GN, among others) has some content available, but I'd love to see more (and if it's worth the hefty $$ 🤯). Thx!
Well, Upsiren PCM-1, Thermalright Helios and PTM 7950 are all the same stuff so ofc they are going to perform the same. Honeywell prints sheets of PTM7950 in 0.2 and 0.25mm in massive rolls, so they could just be buying massive rolls or they are just buying it in bulk as paste and printing the sheets themselves and then rebranding it. At the end of the day, they all cost about the same.
This would definitely make sense. I have no information on whether they are actually all the same product being rebranded, but will let the testing of them be a pretty good decided factor on the value of the products. I haven't seen anywhere to buy the Helios pad yet, but will buy some as soon as I do, or if someone sends me a link to buy it.
I was chatting with El-Hurracan on Reddit and they have a watercooled 3080ti. They did an initial "burn in" up to 65 Celcius. Now it hovers around low 40's to 50 C (25C ambient). They do see idle temps of around 35 C which is higher than what they saw with paste, but they weren't too worried about the idle temps.
Heeyyyy thanks for this, asked you about this test on reddit thanks for getting it out and just in time, just got a tuf 3090 used off ebay for £485 but currently is overheating to 86c on the core, 99 on the vram and 106c on the hot spot!!! Shut it down immediately after seeing that, gpu is fully functional but something is seriously wrong with the cooling solution, tamper sticker was still intact and the seller said he only used it for world of warcraft (overkill i know but i never questioned it) Just hoping with the knowledge you have helped me gain i can get it under control, got upsiren pcm-1some copper shims and ux pro on the way, will post my results after im done Wish me luck lol
That's great to hear you are taking a stab at shims. I would also suggest using either Kapton Tape or Conformal Coating to protect the PCB around the VRAM, just to protect the off chance the shim ever moved (unlikely). I really like the MG Chemicals 422C
Love your test vids! Please do a really good deep dive on grizzly kryosheets. Perhaps not just in cou application but maybe on GPU die too. I was thinking of copper modding and using the kyrosheet in conjunction. But just not sure how safe it'll be. I'm not super worried about placing it in the die but more so worried just about the surrounding area around the vrms
I'll be testing both of these Phase Change Materials out again on Tstbench 2.1 where I log Ambient temperatures in real time within HWINFO64. This will improve the accuracy of the test and be better to use as comparison.
What’s the update? There are so many fakes, and different products to try and figure out… Are the sales on Amazon legit? What about the Kryosheet? What is the best place to try and get genuine PTM7950?
@@warpathh Latest test charts are at this link tinyurl.com/4btsmwu6 Kryosheet does frarily well. Genuine PTM7950 and PTM7958SP can for sure be purchased at Moddiy.
Just a reminder, you put a ton of PCM on the CPU - way more than you needed. It's a costly mistake. They come very thick and it melts to a super-thin layer from 200 microns down to around 20-25 microns, so all the extra thickness that you will have will be simply squeezed out. Best to put like a 20x20mm patch on a 35x35mm IHS, or a 17x17mm patch on a 20x20mm GPU die, because they heat up a lot harder and you shouldn't try to save up too much money on applying PTM7950.
10:10 amazing how this new brand could behave like that, they released a new kind of hybrid between thermal paste and liquid metal I hope you could tested too btw thanks for the extensive testing!
In the case of an AIO, if you only have to melt the pad once, you can do that by turning off the pump and just somehow keeping your CPU temps at 85C. Set a temperature limit of 85C in bios, then hit a heavy benchmark, that'll probably do it. If the pad needs to be in the molten state to perform its best, then it may be worth it to do tests with varying constant pump speeds. Would be pretty funny if 75% pump speed resulted in better temps than 100%
I was chatting woth El-Hurracan on Reddit. They've got a watercooled 3080ti. They did a bit of a burn in up to 65 Celsius, and now it mostly hovers in the low 40's. They did mention that idle temps are a bit warmer than they were with paste. Around 35 Celsius with PTM7950. They seem pretty happy with the overall results.
@AndyU96 so, there's no electrical risk or anything, and they could possibly be used on VRAM. I will likely test it out with shims to see how it goes. One thing I worry about is if the GPU is vertically mounted. Once it reaches phase change temperatures, it's possible the shim may fall out of place and A) no longer be cooling VRAM effectively and B) Make contact with other components, risking a short. I suppose it depends how you'd want to use it.
This advice is so retardedly wrong it hurts, AND it's a good way to fuck up your AIO. Never do this. Disable your power limits instead of disabling the pump. If you do disable the pump there's a high chance you'll boil the coolant in the AIO itself and it being a closed system you really don't want that.
It's one way to do it for sure. lately I've just used room temp with decent success. I now always try to undercut it by 3mm on all sides. Do first peel and lightly rub my finger over to work out any bubble and also to make it stick to the surface. Then I use a small piece of tape or sticker at a corner and do 2nd peel nice and slow. I stop if it pulls up or tears. I usually rub my finger again if that happens, and sometimes try peeling from different corner if need be.
I'm using PTM7950 with a 5950x in a custom loop, and my CPU getting up to 85C with overclocking was enough to burn in PTM7950. So other PCMs ought to work properly as well.
@@snarksdomain Absolutely. With stock settings and an ambient temp of approximately 27.7C, liquid temp of 34.1C, I was able to max out at 75C while gaming. I will be using PTM7950 for all builds going forward, outside of short-term part testing. I even cut off the bits of PTM that were hanging off of the edges of the CPU, and put those scraps on top of the existing PTM on the CPU, effectively double layering it in the middle. My logic was that it would become liquid-ish, and evenly distribute once heated. So it looks like PTM can be "stacked" without negatively impacting performance, as it squishes down during use like it should.
@uss_liberty_incident Awesome. Yeah, if anything it might just take a bit longer to melt and squish out, but the end result should still be a very thin layer of it. They claim as thin as 0.038mm or 0.0015".
Hard for me to be 100% sure, but the packaging was much different, the colour of film is different, they came with different coloured tape tabs, and the little tools included with the Upsiren were not present in the Honeywell packaging.
@prarmageddon yeah, they look similar to me as well. Will try taking better pictures with my camera tomorrow when I get a chance. See if better pics tell us anything different. If they are made by different manufacturers, then it's likely they are using the same thermal booster material. Could be something like Aluminum Oxide, or a similar non electrically conductive Oxide. Heinz ketchup and French's ketchup look the same to me once they are out of the bottle. I prefer the French's myself.
11:19 I picked up some pcm-1 via aliexpress and just got my loop put back together, at 8.5% pump speed (8%=off) running furmark, my 2060 (watercooled it at the height of the cryptocalypse) sitting at 57ºC on the gpu and cpu (5700x) sitting in the low 80s (only running furmark).
@@snarksdomain Didn't get before numbers, was too focused on my loop looking gunky (last time I'll use liquid utopia). A screenshot I have from a couple weeks ago show idle gpu hotspot temp dropped ~6C. Used kryonaut when I assembled the card 1.5-2 years ago, had a dry strip across the gpu core the width of a pencil. Re-running furmark identical to last night I'm seeing ~7C lower on the gpu and ~9C lower on the cpu (ambient temp is roughly the same). Also used CX-H1300 but vrm and ram don't have sensors for me (bought 70g, used about 25g). Applying the pcm wasn't too bad, first application to the gpu was crap (as most first attempts are), but learned to lay one edge down then slowly lay it flat while using a finger on my other hand to press it down, it helped avoid air bubbles, and stuck it in place well enough that removing the second layer of film went much better (started at a corner and film came right up without issue).
hello there hopefully you can answer me, which would it be the best thermal pad to use from aliexpress for cpu and gpu ? and if you also tell the thickness that would be helpful, for reference ryzen 7800x3d and rtx 4070
From aliexpress for the CPU and GPU core I'd either go Honeywell PTM7950 sheet, Thermalright Heilos, or Upsiren PCM-1. They are phase change and melt at 45 Celsius and squish down to 0.038mm (0.0015"). For everywhere else that uses thermal pads, like VRAM and VRM, I'd use a Thermal Putty instead such as Upsiren UTP-8 or Fehonda LTP81
In my quest of finding phase change silicone pads, I came across this new one. Have you heard of the Maxtor XTP-001? They claim 22.5w/mk thermal conductivity. Which I find strange, since it's so much higher than the Upsiren and Honeywell. I can't find any solid reviews for it. Just ali express listings and other pages in online shopping platforms. It would be interesting if you can get your hands on one just to see if it's better or not.
@Vitorlean From this test I would conclude that they perform the same. I'll be retesting them on Testbench 2.0 and logging ambient within HWINFO64 to get even more accurate results. On this test I was still using the standalone temperature sensors, which aren't logged in real time.
@Vitorlean ModDIY is an official reseller of Honeywrll PTM7950 so it is the guaranteed original PTM7950. I'm not sure if that answers your question or not.
I can definitely do that once I get it figured out. Amazon canceled my order, so I reordered the Quadro from Performance PC last night. I received 2 XSPC 10K sensors yesterday though :)
Update: Okay, so I didn't really have to do anything for it to show up in HWINFO64. It was there waiting for me after I installed the Aquasuite software and rebooted. Äquacomputer Quadro"showed up in it's own drop down menu.
Once these melted and disassembled the heatsink, do you have to replace them before putting the heatsink back? Just check the price on the sheet and they're pretty pricey
Having taken off the heatsink several times during the testing to replace it, I could see it was a nice even layer each time. I think you could try putting the cooler right back on and see what it does. If you had any tiny pieces of a sheet left over, you could even put a tiny piece in the middle of the CPU. As it melts, it should flow out and help push any potential new air bubbles out.
I wouldn't put too much weight/importance on the advertised w/mk rating of any Thermal Interface material. The ratings are used to help market a product. De8auer and Gamers Nexus did a really good video where they talked about the issue. I've even seen 2 products from the same company have a higher and lower rated product. And the one with the lower w/mk rating outperformed the one with the higher rating, significantly. These days I only trust testing. I do use advertised ratings as a rough guide on which products to test within a product lineup though. I'd also add that you can find PTM7950 for similar price. I've tested some from a few different sources and they all seem to perform fairly well.
I still need to get to testing that along with a bunch if other paste. PCM, and liquid metal solutions. Just working through retesting the putties currently for a few more days and then will start on that.
They are the same for ease of application. I now cut the piece smaller than the die/IHS by about 3mm (1/8") around each side. It makes it easier to peel I find. I peel one side of protective film, then I place the piece on the Die/IHS and rub my finger on it to make sure it's sticking well. If I see any large bubble I press them out. Then I use a little piece of tape on a corner and peel the 2nd protective film off (packing tape works well).
Good video, thanks. Would you happen to know if pcm-1 has similar longevity to 7950 in real world applications. I've seen many instances of people saying 7950 doesnt degrade well after a year (if it does it seems to be within margin of error). Thanks again for the video, you're one of the few covering thermal putties and "niche" products like this.
Yes I did. I've since retested several different Phase Change Materials from a few different sources. I've found it is difficult to differentiate PCM/Paste performance with my current setup, even with tracking Ambient temps (only ~3 degree span from best to worst) Once I make some further improvements to my test bench and test methodology I'll go back and retest everything again. Currently focused on testing Thermal Pads, Putties and Shims. This is much easier to do with 35 degrees from best to worst, even considering the 2 degree resolution on Nvidia Mempry Junction Temperaturws within HWINFO.
@@snarksdomain Thanks for the answer :) I bought honeywell from aliexpress and moddiy a few months ago and it came with different foils, PTM from aliexpress came with grey foil and PTM from moddiy came with blue foil, their consistency was also different, PTM from aliexpress was easier to put on, to put PTM from moddiy I had to put it in fridge for 30 minutes first :D I came to the conclusion that all PTM from aliexpress are fake, they feel more like Upsiren than original Honeywell but they perform almost identical at least for the first few months, what is your opinion about this? Do you also have PTM from aliexpress to compare? Regards :)
@@snarksdomain Thanks for the answer :) I bought honeywell from aliexpress and moddiy a few months ago and it came with different foils, PTM from aliexpress came with grey foil and PTM from moddiy came with blue foil, their consistency was also different, PTM from aliexpress was easier to put on, to put PTM from moddiy I had to put it in fridge for 30 minutes first :D I came to the conclusion that all PTM from aliexpress are fake, they feel more like Upsiren than original Honeywell but they perform almost identical at least for the first few months, what is your opinion about this? Do you also have PTM from aliexpress to compare? Regards :)
It seems that way at first glance of the charts, and it's possible that PTM7950 outperforms PCM-1. With my current method of testing I believe I am out +/- 0.5 degrees Ceclsius, or around that. You will see results jump around between the CPU tests and the GPU Core tests. I'm hoping to make my testing more accurate in the future. There isn't enough of a spread with Pastes and Phase Change Materials results to be confident on the findings for me. With Thermal Pads and Putties there is a much larger difference in results so it makes it easier to see a difference in performance. Even still, pad/putty results jumo around between ETCHASH, Furmark, and KAWPOW results. I'm starting to form how I want to set up Testbench 3.0 in my mind. Will likely wait for RTX 50 series so we can test GDDR7 and see how hot it gets.
@@snarksdomain your test are good enough:) really great work. After checking Your charts I decided to go for CX H1300 13.5W instead of upsiren u6 pro, and ptm7950 instead of pcm-1 based on Your latest charts. I can use ptm for gpu and cpu if I am not mistaken right? Cheers
@@snarksdomain You do a great job. Based on Your results I will pick PTM7950 over PCM-1, and CX-H1300 over UPSIREN U6 PRO. That make sense do You think? BTW I can use PTM7950 for both CPU and GPU right? cheers
@smokehead5318 For sure. You can definitely use PTM7950 on both. I recommend giving it a good burn in for 2 hours. If you can get CPU temp up to around 65-70 Celsius if should get a good melt and become nice and thin. I saw the largest improvements in the first 2 hours for most of the PCM's, the exception being Laird TPCM7250, which needed a hotter burn in to get a proper melt (I'd recommend 85-95 Celsius burn in for Laird).
@@snarksdomain ok thank You for a such informative reply. Cx-h1300 is superior to u6 pro based on Your test, however it’s a bit stiffer during application?
For sure. Once I get through all these thermal putties I'll be testing out Upsiten LMTG-100 and then Thermal Grizzly Conductonaut. After those I'll do some thermal pastes (all the ones I have on hand), and then the Phase Change Materials. PCM's will take the longest to test.
@@snarksdomain thankyou sire! They claim there wont be "leak" like your typical lm (or thats what i read lol)..... Also speaking of pcm.... Liano also made a pcm that look identical to upsiren(the ads rofl) but appereanly it has "up to 13.8w/mk" might be worth a try lol
Phase Change Material. Either Honeywell PTM7950, Upsiren PCM-1, Thermalright Heilos, or even Laird TPCM7200 (requires higher temperature to melt than the other 3)
@edwardscroop9149 you can. These days I just use it a room temp, and cut it smaller by 3mm on all 4 sides. Makes application super easy, and 2nd peel no issue.
@@snarksdomain Can you please retest Copper shims? You only have tested them in the obsoelete chart, i want to see whether it's best to keep the current 3d graphite pad on my smartphone soc or replace with copper shim and paste?? Your videos are actually helping alot, thank you for your work.
@hotsauce6310 I'm testing out Aluminum Nitride shins and Putty roght now and will test the other shim and putty combos soon. One nice thing about the AIN shims is that they don't conduct electricity.
@@snarksdomain I'm not really sure where to find those. I did not find graphene pads to cause any harm yet, i guess its the little project of mine, seeing it's longevity and it's reliability on my older phone. Seems to perform better than i remember and compared to others. Not sure by how much, note that i had to build the phone from scratch after 4 years of no use, maybe you could drop some insight on it? what about your obsoelete chart? There the graphite pads are no better than thermal putty. Wrong methodology used? or Pads size? I've seen u have to use higher than actual thermal pad size because graphite pads performs better when compressed. In my case i took off some top shielding that m/b had over soc & some other components and did a direct IC(Ram stacked on top of Cpu) to frame/housing application.
@@snarksdomain I saw somewhere that graphite pads only last around 6 months? is that true? What happens to them after it they get dry or does thermal conductivity drop significantly?
Hi guys! I want to make a really low power setup out of a ryzen 4750ge I got my hands on a few weeks ago. Would this get hot enough to get the ptm to melt?
40x80 is definitely enough. If intel CPU, you'll likely be using 30x40mm. That would leave you 50x40mm, which is lots. You might even b3 able to mess up the CPU application once and still have enough for GPU. If AMD then you'll be using closer to 40x40mm and will still have lots for GPU, but try your best to get the CPU application correct the first time. Likely best to do CPU first and if it messes up you might be able to salvage a big enough piece for GPU. Even if you have to use a few broken pieces, I would still go through with it. This stuff will melt and make a nice even layer in the end, as long as you have enough to melt and spread.
Way more than enough. You don't need nearly as much PTM7950 as you think. They come very thick and it melts to a super-thin layer from 200 microns down to around 20-25 microns, so all the extra thickness (volume) that you will have will be simply squeezed out. Best to put like a 20x20mm patch on a 35x35mm IHS, or a 17x17mm patch on a 20x20mm GPU die, because they heat up a lot harder and you shouldn't try to save up too much money on applying PTM7950. If it's a 35x42mm IHS like with Intel LGA1700 processors, just use a 20x24mm patch.
@@one_step_sideways THis is a good point. I'used large squares simply because the smaller sheets I got aren't large enough to have enough excess for other application. I do have 1 large sheet of PTM7950 that I bought off Aliexpress and need to test to see if it's is real, but it looks real from the packaging and labelling. I'll definitely be cutting smaller pieces from sheet.
@@snarksdomain I wonder if you can put something like 4 small patches that are like 8x8mm and place them 5-10mm apart to guarantee the best spread without using too much paste by simulating an X-pattern like with thermal paste, since the spreader method is expensive. Because with a 20x20mm patch in the middle you have the same kind of spread as with a drop of paste, and it may still be too much, considering the 5-10x reduction in thickness. I think you should test that with your sheet of cheap PTM7950 you got. You can put the 8x8mm patches on the IHS like so: ____________ | | | × × | | | | × × | |_________|
Tôi thực sự là fan hâm mộ của bạn. Tôi cũng đang điều hành một doanh nghiệp dịch vụ vệ sinh máy tính xách tay tại Việt Nam. Tôi có thể liên hệ với bạn ở đâu để hỏi một số câu hỏi liên quan đến mod kim loại lỏng và đồng vram
You can contact me at snarksdomain@gmail.com for questions if you like. I don't have experience with Liquid Metal yet, so will be studying the subject more before I go to test it.
Using phase change material on a CPU is a waste, even for testing purposes. Despite CPU using about 150W of power, this does not translate into 150W of radiated heat. All new Ryzen CPUs, be it 7nm chip models or newer, have very small die size, which limits heat output. My personal guess, about 80W gets out - that´s why with these CPUs, low-end 120mm tower coolers are able to compete with high-end 140mm dual towers. Even intel is on the same boat now, where latest gen chip is so dense, despite consuming lot of power and running hot, can´t be cooled down with anything. Also, even small changes in ambient affect the results. I´d steer away from the Upsiren too. It´s an unknown brand with unknown history. The product claims to be related to "computer-systems gr" printed on the box, which is actually k5 Pro putty manufacturer and they themselves distanced from Upsiren product brand. Even when it seems to work just fine, there is no documentation for... anything. Chinese sellers also took your data and are now claiming "our products have been tested", aka they themselves had no idea, how good/bad the stuff actually is.
I think it's worth testing Phase Change Materials, at least I'm having fun doing so. I'll be testing more accurately with ambient temperature being logged using an Aquacomputer Quadro. Upsiren is a good brand from all my testing so far. I agree, they are not connected to Computer-Systems.gr, thankfully. I'm not entirely sure why Upsiren listed that website on their packaging. I assume it was an attempt to help break into a competitive market of Thermal Interface Materials. That being said, I think they can stand on their own two legs now, and should likely remove said website from their packaging, to make things less muddy. It's good to remember that most Thermal Interface Materials are manufactured in countries such as China, Malaysia, Vietnam, and South Korea. It is not unusual to have new companies start to create new materials for an ever growing market. From talks with Upsiren, they have done their own internal testing of the products they manufacture. and there are some new products coming that I will be testing out as well. That being said, I'm also looking forward to testing out more products from Fujipoly, Laird, Mg Chemicals, Thermal Grizzly, and more.
@@snarksdomain it´s usually a tactic to confuse customers into thinking, they´re buying stuff made by renown manufacturer. A good quality product can speak volumes by itself and does not need aggressive (fake) advertising. It´s not surprising, everything is manufactured in East Asia. What is concering, however, is the total lack of documentation. What are the properties? Temperature for use? Temperature for storage purpose? etc.
@Morpheus-pt3wq Yeah, it would be better if Upsiren removes the Computer Systems website from the packaging. Other than the basic specs, it would be nice to see some more documentation for sure. I'll be doing some longevity testing of a several different products but won't have those results for at least a year (perhaps I should do confirmation testing in 1 year intervals and see how the products hold up.).
@Morpheus-pt3wq All of the heat the CPU generates must get out. 150W in = 150W out. It just takes a larger temperature difference between the die and cooler when the die is smaller. Combine that with the insane power consumption of the highest end models and you have trouble cooling them. These testing vids are a great. Once the data logging gets improved, even better. No one else compared so many brands of putty and pads.
For some reason your comment was stuck in "Held for Review", which is not something I often check. My appologies for not responding until now. I will look up that LLANO Phase Change and see if I can get some for testing. Thanks for the suggestion.
I don't normally interact with these videos, but I really appreciate the effort you put into this. Please test the thermalright solution when you can! Also, maybe try cutting the squares into a more accurate size, or ever ever so slightly less than the full coverage as it still spreads out. Should have less overspill and may result in better temps.
Thank you for your comment. That's a good idea on undercutting it a bit. Will definitely try that out. I used to do that with pastes, where I would spread it and then wipe a 1/8" (3mm) border clean of paste and still have great contact patch.
Use a plastic spatula or guitar pick to drag along the edges of the die or IHS where the PTM overhangs, it will cut the PTM and make it super easy to peel the film without lifting or tearing.
I still don't think it matters at all if there are some air bubbles, since they should just squeeze out once there's heat and pressure. DerBaur has also stated as much, if you don't believe me.
dude thanks, I was figuring out similar idea U shape but start from one edge. Mine pad should arrive in couple weeks and I try your style first :) I got an MSI p65 laptop. CPU is power limited but I hope I can drop temps and by that drop the fan noise. If it cools my laptop I slap it also on my desktop that probably is due for thermal paste replacement as its been running since ryzen 1700 release without any re-paste :D still has the stock thermal paste pre applied on the wraith cooler so I imagine huge cooling boost :p
I gave you thumbs up, as it's pretty smart application here, BUT...air bubbles will be ironed out by immense heat generated within the cpu, and it will do it darn quickly. Similar like protective glass bubbles on the phone go away after normal usage, and that screen barely gets warm.
I would say this is a safe bet to stop being paranoid over some bubbles. We are good. Just the pressure of the cpu mounting/waterblock + heat, and you don't have to do summersaults around the application methods.
That's my experience. I've put on protective films on glass as well, and heat does its' work.
@@righteousone8454 Yeah you're right, bubbles don't make any difference once it melts. I edited the comment to remove the superfluous stuff.
Nope, tried everything, did this literally yesterday and it was my second attempt at this. I out it in the freezer for a while to make it easier too. Nope, still teared in several places. It's like it's glued to the foil. Of the 50EUR I paid for it, I maaaaybe have enough left for another application. It was a nightmare and I really am not an graceless elephant when it comes to stuff like that.
I had to patch it up, but then it was so uneven and thick in some places, that it was locking my PC when I wanted to run Cinebench for example. I had to take the cooler off, scrape the excess PTM, put it on again and now it's working as intended. But man... it's the worst.
It's cool that more companies are releasing PCM. I know that there is still Laird Tpcm 7125 7,5 W/mk. I wonder if Thermalright Heilos 8,5 W/mk will also be PCM, but so far I have not seen such information.
Thanks for the test 👍
I'm hoping to buy some Helios soon, but haven't found an online retailer yet. Will check out that Laird PCM when I can.
Edit: In my opinion it likely is, but only time will tell.
Laird hasn't responded back to me yet, so I ordered some off Digikey. I went with the Tpcm 7250 (A18172-10).
Looking forward to your tests on Thermalright‘s Heilos pad 🤓👍
As soon as I can buy it I will :)
@@snarksdomain You should do a shootout with PCM1, PTM7950, The LTT pad, Thermalright, etc.
@@jamesdavies686 I had compared 3 of those on Testbench 2.1 but decided to stop core/CPU testing on that testbench as I didn't feel it was accurate enough to trust the results, with such a small temperature span.
I'll soon be testing all of these things, and more PCM's on Testbench 3. I don't plan to buy the PTM7950 from LTT as it will be the same as the one from Moddiy. I also have some sheets coming from Caplinq. I do have Gelid, Thermal Grizzly and Maxtor PCM's though, among others.
Testbench 2.1 will continuue with Putty/Pad/Shim testing for now so we can fill out the chart a bit more (mostly shim+putty combos).
All the current charts are located at the Google Drive link below (Test Chart Repository)
Great and decent info man, without the extra misleading stuff other channels do/add, keep up the good work, and keep improving your channel, you have a decent informational channel here that I would trust more than the big channels. Good luck.
Hey there, with PTM thermal pad and my experience I found that sticking it in the refrigerator for around a half hour or so makes application a hell of a lot easier. Idk what your house ambient temp is, but cooling it in the fridge definitely helped me out. It made peeling the plastic applicator extremely easy and the pad remained whole.
This is a solid tip for sure. I've done the fridge technique a couple times with good success.
Some Phase Change sheets now have overhanging plastic films on both side which makes first peel super easy. And second one just as easy.
For the sheets without the extra plastic film, those little sticker tabs certain help. Can put one on either side of plastic at one corner and peel apart without much fuss.
Another cool trick, and this one from T.Sharp, is using a credit card or other firm plastic holding it at a 45 degree angle and running it along the edge of the IHS/Die with modest pressure to "chop" off any excess. This makes second peel go smooth and doesn't lift a flap at the corner.
Several others have also suggested undercutting the square by a couple mm on each side as this will use less of a large sheet and give more applications. This won't be a huge benefit for small sheets though (where it's only large enough for 2 CPU applications.
Nope, tried everything, did this literally yesterday and it was my second attempt at this. I out it in the freezer for a while to make it easier too. Nope, still teared in several places. It's like it's glued to the foil. Of the 50EUR I paid for it, I maaaaybe have enough left for another application. It was a nightmare and I really am not an graceless elephant when it comes to stuff like that.
I had to patch it up, but then it was so uneven and thick in some places, that it was locking my PC when I wanted to run Cinebench for example. I had to take the cooler off, scrape the excess PTM, put it on again and now it's working as intended. But man... it's the worst.
@@EpicKillstreak is it tearing when you peel the first plastic film? Or is it tearing after you've pushed it down onto the CPU and ran your finger over it, and the peeling second film?
The ladder, after its pushed to the cpu
@EpicKillstreak that's no fun. It can be tricky to get it to peel. I had the best luck with the 0.25mm phase change sheets, and T.Sharps method of running a plastic spatula around the edge at a 45 to kind of cut it.
I tested the upsiren pcm 0,02mm vs the ptm7950 from aliexpress (0,25mm), specifically from passionate girl store and this was a little better than the upsiren. The only downside with these pads is when you take them out (specially on laptops because of the small bare dies) after some months I made a rookie mistake and ended up cracking the cpu die on my msi laptop because i tried to remove the heatpipe when the laptop was cold, so it's better to warm up everything if you need to dissasembly your laptop. Confirmed this a couple of days ago, used some pcm on copper sheets, warm'em up and then let them cool down and dude it was hard af to separate them without heat. Anyway great channel, love your videos, a lot of information we don't find anywhere else, thanks!
Phase Change materials definitely become hard once they cool down (any temperature below 45C of the material itself will be solid).
If anyone uses Phase Change on an AM3 or AM4 CPU they will need to use the AMD recommended "Twist" method if/when removing the cooler. This goes for dried paste as well.
interesting insight, will definitely keep this in mind once I repaste mine with PTM
Before i replace thermal paste with PTM my theory is a D-Limonene cleanup (Arcticlean is an option, step 1 bottle is D-Limonene i think) followed up with 99% alcohol thorough clean for the silicone residue from the old paste increases the adhesion and stability, esp inside the micro-pores from the cold-plate, dabbed on alcohol must lay flat on the surfaces without moving away or forming droplets, that usually means the surface is still hydrophobic from silicone residue. On heatsinks i use a small gas torch (comparable with a torch cigarette lighter) with short bursts, it oozes out the remaining silicone or whatever that liquid is inside the pores, wipe off with a piece of paper shop towel. Done this on 100's of heatsinks. After that the new material fills up those empty voids without silicone interference.
On car spraypainting silicones are really nasty stuff, 1 droplet of silicone on a panel (tire shine and such) and the respray is trash.
That's a great tip. I've used a little butane lighter a few times, and it's surprising to see how much residue oozes out. I will look into D-Limonene, thanks for the tip.
@@snarksdomain No need to go with pure d-limonene.
Just go with something like a citrus degreaser and it'll clean everything :)
A fun experiment to try and would be interesting to see the results.
Instead of using the usual thermal pads on Vram/Ram chips, can you experiment with using PTM or PCM instead (even double layer it) and see if it gets good results.
It sounds like a cool idea, but that would be stacking up to achieve 1.8mm gap (so about 8-9 layers of PCM material. If/when it melts, it would then basically glue the VRAM to the Cooler and would be extremely difficult to pull apart. This would also be more expensive than putty or pads.
I mostly just use a high end Thermal Putty on VRAM/VRM these days, rather than pads.
I might try shims and PCM on a card I'd be willing to sacrifice someday. Just worried about damage when trying to pull it back apart.
@@snarksdomain yeah it would be cool to experiment on an oIder card you don't care about anymore, with either shims or PCM.
Definitely very curious to see what happens.
Thermalright released similar products too, the name is Heilos. Maybe this can add into the comparison in the future. Nice video as always.
Got some Helios in hand for testing :)
@@snarksdomain nice
Nah Helios kinda sucks compared to ptm7950
@@godnamedtay both are PTMs, what are you even on about?
Just found your channel (obviously new to me...lol). Your methodology seems quite thorough, so kudos for that. Looking forward to your future installments! Keep up the great work.
Thank you, and welcome to the channel. Lots of fun stuff to come. I'll be testing out some different viscosities of the Upsiren Putties soon too. Will likely start that testing once I'm back home.
@@snarksdomain Awesome, very interested in your results. Have you done any work with the Thermal Grizzly KryoSheet? Roman (and GN, among others) has some content available, but I'd love to see more (and if it's worth the hefty $$ 🤯). Thx!
@@danytoob I have some Kryosheet and will be testing it once I get back to testing pastes and PCM's.
@@snarksdomain 👍👏🖖
Well, Upsiren PCM-1, Thermalright Helios and PTM 7950 are all the same stuff so ofc they are going to perform the same.
Honeywell prints sheets of PTM7950 in 0.2 and 0.25mm in massive rolls, so they could just be buying massive rolls or they are just buying it in bulk as paste and printing the sheets themselves and then rebranding it.
At the end of the day, they all cost about the same.
This would definitely make sense. I have no information on whether they are actually all the same product being rebranded, but will let the testing of them be a pretty good decided factor on the value of the products. I haven't seen anywhere to buy the Helios pad yet, but will buy some as soon as I do, or if someone sends me a link to buy it.
for Watercooling getting it hot shouldn't be a problem, can just throttle the Pump or even turn it off for a few Minutes and it'll get nice and warm.
I was chatting with El-Hurracan on Reddit and they have a watercooled 3080ti. They did an initial "burn in" up to 65 Celcius. Now it hovers around low 40's to 50 C (25C ambient). They do see idle temps of around 35 C which is higher than what they saw with paste, but they weren't too worried about the idle temps.
Heeyyyy thanks for this, asked you about this test on reddit thanks for getting it out and just in time, just got a tuf 3090 used off ebay for £485 but currently is overheating to 86c on the core, 99 on the vram and 106c on the hot spot!!! Shut it down immediately after seeing that, gpu is fully functional but something is seriously wrong with the cooling solution, tamper sticker was still intact and the seller said he only used it for world of warcraft (overkill i know but i never questioned it)
Just hoping with the knowledge you have helped me gain i can get it under control, got upsiren pcm-1some copper shims and ux pro on the way, will post my results after im done
Wish me luck lol
That's great to hear you are taking a stab at shims. I would also suggest using either Kapton Tape or Conformal Coating to protect the PCB around the VRAM, just to protect the off chance the shim ever moved (unlikely).
I really like the MG Chemicals 422C
@snarksdomain3453 yeah got some kapton tape which should be sufficient I hope ^^
How did it go?
Love your test vids! Please do a really good deep dive on grizzly kryosheets. Perhaps not just in cou application but maybe on GPU die too. I was thinking of copper modding and using the kyrosheet in conjunction. But just not sure how safe it'll be. I'm not super worried about placing it in the die but more so worried just about the surrounding area around the vrms
I'll be testing both of these Phase Change Materials out again on Tstbench 2.1 where I log Ambient temperatures in real time within HWINFO64. This will improve the accuracy of the test and be better to use as comparison.
What’s the update? There are so many fakes, and different products to try and figure out… Are the sales on Amazon legit? What about the Kryosheet? What is the best place to try and get genuine PTM7950?
@@warpathh Latest test charts are at this link tinyurl.com/4btsmwu6
Kryosheet does frarily well. Genuine PTM7950 and PTM7958SP can for sure be purchased at Moddiy.
Just a reminder, you put a ton of PCM on the CPU - way more than you needed. It's a costly mistake. They come very thick and it melts to a super-thin layer from 200 microns down to around 20-25 microns, so all the extra thickness that you will have will be simply squeezed out. Best to put like a 20x20mm patch on a 35x35mm IHS, or a 17x17mm patch on a 20x20mm GPU die, because they heat up a lot harder and you shouldn't try to save up too much money on applying PTM7950.
he's a noob
I got my ptm Honeywell for free so I personally don't care lol
looking foward to your test on helios pad :P and probally ptm 7958 lol
I haven't seen the Helios pad for sale on any websites/e-tailers yet. Not sure when it will be available for purchase. Hopefully soon.
10:10 amazing how this new brand could behave like that, they released a new kind of hybrid between thermal paste and liquid metal I hope you could tested too btw thanks for the extensive testing!
Hi, can you test the new product from Upsiren MTG-100 liquid metal? I want to see the results of this liquid metal
Upsiren UTG-X seems awesome, it had the highest ambient temps to deal with
I plan to test it more. Seems like a decent paste. I've been focusing so much on Phase Change Materials lately
In the case of an AIO, if you only have to melt the pad once, you can do that by turning off the pump and just somehow keeping your CPU temps at 85C. Set a temperature limit of 85C in bios, then hit a heavy benchmark, that'll probably do it.
If the pad needs to be in the molten state to perform its best, then it may be worth it to do tests with varying constant pump speeds. Would be pretty funny if 75% pump speed resulted in better temps than 100%
I was chatting woth El-Hurracan on Reddit. They've got a watercooled 3080ti. They did a bit of a burn in up to 65 Celsius, and now it mostly hovers in the low 40's.
They did mention that idle temps are a bit warmer than they were with paste. Around 35 Celsius with PTM7950. They seem pretty happy with the overall results.
@@snarksdomainCan these phase change pads only be used with core dies or can they also be used on the VRAM modules as well?
@AndyU96 so, there's no electrical risk or anything, and they could possibly be used on VRAM. I will likely test it out with shims to see how it goes.
One thing I worry about is if the GPU is vertically mounted. Once it reaches phase change temperatures, it's possible the shim may fall out of place and A) no longer be cooling VRAM effectively and B) Make contact with other components, risking a short.
I suppose it depends how you'd want to use it.
This advice is so retardedly wrong it hurts, AND it's a good way to fuck up your AIO.
Never do this. Disable your power limits instead of disabling the pump. If you do disable the pump there's a high chance you'll boil the coolant in the AIO itself and it being a closed system you really don't want that.
Phase change materials need to be cooled in refrigerator before install. Better for cutting and installing!
It's one way to do it for sure. lately I've just used room temp with decent success.
I now always try to undercut it by 3mm on all sides. Do first peel and lightly rub my finger over to work out any bubble and also to make it stick to the surface. Then I use a small piece of tape or sticker at a corner and do 2nd peel nice and slow. I stop if it pulls up or tears. I usually rub my finger again if that happens, and sometimes try peeling from different corner if need be.
Can you test the TG A7000 thermal putty? Curious about the comparisons to the other putties.
I'm using PTM7950 with a 5950x in a custom loop, and my CPU getting up to 85C with overclocking was enough to burn in PTM7950. So other PCMs ought to work properly as well.
Nice. Did you end up with decent temperatures after the burn in?
@@snarksdomain Absolutely. With stock settings and an ambient temp of approximately 27.7C, liquid temp of 34.1C, I was able to max out at 75C while gaming. I will be using PTM7950 for all builds going forward, outside of short-term part testing.
I even cut off the bits of PTM that were hanging off of the edges of the CPU, and put those scraps on top of the existing PTM on the CPU, effectively double layering it in the middle. My logic was that it would become liquid-ish, and evenly distribute once heated. So it looks like PTM can be "stacked" without negatively impacting performance, as it squishes down during use like it should.
@uss_liberty_incident Awesome. Yeah, if anything it might just take a bit longer to melt and squish out, but the end result should still be a very thin layer of it. They claim as thin as 0.038mm or 0.0015".
Aren't they the same product but just branded differently?
Hard for me to be 100% sure, but the packaging was much different, the colour of film is different, they came with different coloured tape tabs, and the little tools included with the Upsiren were not present in the Honeywell packaging.
@@snarksdomain but the substance itself, does it look the same? It is hard to tell by the video.
@@prarmageddon Here is a comparison using the camera on my phone: imgur.com/a/6MwT8xP
@@snarksdomain thank you, they look identical to my eye
@prarmageddon yeah, they look similar to me as well. Will try taking better pictures with my camera tomorrow when I get a chance. See if better pics tell us anything different.
If they are made by different manufacturers, then it's likely they are using the same thermal booster material. Could be something like Aluminum Oxide, or a similar non electrically conductive Oxide.
Heinz ketchup and French's ketchup look the same to me once they are out of the bottle. I prefer the French's myself.
Thank you for this tedt
bro, pls test PTM7950 OG vs new PTM7958.
11:19 I picked up some pcm-1 via aliexpress and just got my loop put back together, at 8.5% pump speed (8%=off) running furmark, my 2060 (watercooled it at the height of the cryptocalypse) sitting at 57ºC on the gpu and cpu (5700x) sitting in the low 80s (only running furmark).
Let us know how your temperatures look in a couple days and if it has been an improvement for you.
@@snarksdomain Didn't get before numbers, was too focused on my loop looking gunky (last time I'll use liquid utopia). A screenshot I have from a couple weeks ago show idle gpu hotspot temp dropped ~6C. Used kryonaut when I assembled the card 1.5-2 years ago, had a dry strip across the gpu core the width of a pencil.
Re-running furmark identical to last night I'm seeing ~7C lower on the gpu and ~9C lower on the cpu (ambient temp is roughly the same). Also used CX-H1300 but vrm and ram don't have sensors for me (bought 70g, used about 25g).
Applying the pcm wasn't too bad, first application to the gpu was crap (as most first attempts are), but learned to lay one edge down then slowly lay it flat while using a finger on my other hand to press it down, it helped avoid air bubbles, and stuck it in place well enough that removing the second layer of film went much better (started at a corner and film came right up without issue).
Please test these on the GPU Die
hello there hopefully you can answer me, which would it be the best thermal pad to use from aliexpress for cpu and gpu ? and if you also tell the thickness that would be helpful, for reference ryzen 7800x3d and rtx 4070
From aliexpress for the CPU and GPU core I'd either go Honeywell PTM7950 sheet, Thermalright Heilos, or Upsiren PCM-1. They are phase change and melt at 45 Celsius and squish down to 0.038mm (0.0015").
For everywhere else that uses thermal pads, like VRAM and VRM, I'd use a Thermal Putty instead such as Upsiren UTP-8 or Fehonda LTP81
In my quest of finding phase change silicone pads, I came across this new one. Have you heard of the Maxtor XTP-001? They claim 22.5w/mk thermal conductivity. Which I find strange, since it's so much higher than the Upsiren and Honeywell. I can't find any solid reviews for it. Just ali express listings and other pages in online shopping platforms. It would be interesting if you can get your hands on one just to see if it's better or not.
Thjanks for the heads up! I'll take a look
@@snarksdomain No problem! Looking forward to it.
@@conyo985 I've got 4 on the way. 2 for CPU and 2 for GPU.
@@snarksdomainGreat news! I'm getting excited!
Thanks for the videos really helpful! Have you tested Thermalright Heilos?
Indeed I have. Test charts are at the link in the channel about page :)
@@snarksdomain thank you bro
What is best upsiren pcm 1 or ptm7950 of Aliexpress?
I still need to fully test the Aliexpress PTM7950.
@@snarksdomainupsiren pcm 1 have similar performance of ptm 7950 original? Or ptm 7950 original is very best?
@Vitorlean From this test I would conclude that they perform the same. I'll be retesting them on Testbench 2.0 and logging ambient within HWINFO64 to get even more accurate results. On this test I was still using the standalone temperature sensors, which aren't logged in real time.
@@snarksdomain but this test was with ptm 7950 of moddiy? So the chineses know very do thermal paste.
@Vitorlean ModDIY is an official reseller of Honeywrll PTM7950 so it is the guaranteed original PTM7950. I'm not sure if that answers your question or not.
Can you please do a guide on the temp sensor adding to HWinfo64? That would be neat!
I can definitely do that once I get it figured out. Amazon canceled my order, so I reordered the Quadro from Performance PC last night. I received 2 XSPC 10K sensors yesterday though :)
Update: Okay, so I didn't really have to do anything for it to show up in HWINFO64. It was there waiting for me after I installed the Aquasuite software and rebooted. Äquacomputer Quadro"showed up in it's own drop down menu.
Once these melted and disassembled the heatsink, do you have to replace them before putting the heatsink back? Just check the price on the sheet and they're pretty pricey
Having taken off the heatsink several times during the testing to replace it, I could see it was a nice even layer each time. I think you could try putting the cooler right back on and see what it does.
If you had any tiny pieces of a sheet left over, you could even put a tiny piece in the middle of the CPU. As it melts, it should flow out and help push any potential new air bubbles out.
It practically has the same thermal conductivity and Upsiren takes it up or down a notch by being cheaper Honeywell after all is very old company
I wouldn't put too much weight/importance on the advertised w/mk rating of any Thermal Interface material. The ratings are used to help market a product.
De8auer and Gamers Nexus did a really good video where they talked about the issue.
I've even seen 2 products from the same company have a higher and lower rated product. And the one with the lower w/mk rating outperformed the one with the higher rating, significantly. These days I only trust testing. I do use advertised ratings as a rough guide on which products to test within a product lineup though.
I'd also add that you can find PTM7950 for similar price. I've tested some from a few different sources and they all seem to perform fairly well.
im sorry if you have already done it but are you planning on comparing the cryosheet to these other solutions?
I still need to get to testing that along with a bunch if other paste. PCM, and liquid metal solutions.
Just working through retesting the putties currently for a few more days and then will start on that.
When applying the PTM7950/PCM-1, do you trim the edges so that it doesn't stick out over the IHS' edge before putting on the cooler?
You could. I haven't been.
@snarksdomain3453 what about on the GPU? I was thinking of applying it on the GPU processor as well as my CPUs IHS.
@DivineDragonCrisis Yes, you can also use it on the GPU die as well. This stuff doesn't conduct electricity.
@@snarksdomain alright. I'll probably put the PTM7950 or PCM-1 on the CPUs IHS and the GPU die, and puttys on wherever else.
is one easier to apply than the other or are they about the same?
They are the same for ease of application.
I now cut the piece smaller than the die/IHS by about 3mm (1/8") around each side. It makes it easier to peel I find.
I peel one side of protective film, then I place the piece on the Die/IHS and rub my finger on it to make sure it's sticking well. If I see any large bubble I press them out. Then I use a little piece of tape on a corner and peel the 2nd protective film off (packing tape works well).
have you heard of thermalright helios? they say it is ptm7850 @@snarksdomain
What about upsiren utg-x thermal paste. They says 14.8 w/mk?
It's also in the chart in this video. Forgot to snap before pictures of applying it. It's currently on the CPU of the test bench.
I have half of the 8g tube of Arctic MX-2 . Works perfectly on my gaming laptop. Change it once per 1.5 year. Will not trust these weirdo pads :)))
Totally fair. You gotta go with whatever you're comfortable with.
Good video, thanks.
Would you happen to know if pcm-1 has similar longevity to 7950 in real world applications. I've seen many instances of people saying 7950 doesnt degrade well after a year (if it does it seems to be within margin of error).
Thanks again for the video, you're one of the few covering thermal putties and "niche" products like this.
We won't know that for quite some time, unfortunately. I'll have to see over time how its performance holds up.
@@snarksdomain 100% understandable, thanks for the reply, appreciate it!
PTM7950 lasts 3-5 years in industry, so in a household system 10+ years, since PCM is the same stuff so it will last a while
@pokehl9929 That would be super cool if it lasts that long or longer.
Hi, did you get honeywell from moddiy?
Yes I did. I've since retested several different Phase Change Materials from a few different sources.
I've found it is difficult to differentiate PCM/Paste performance with my current setup, even with tracking Ambient temps (only ~3 degree span from best to worst)
Once I make some further improvements to my test bench and test methodology I'll go back and retest everything again.
Currently focused on testing Thermal Pads, Putties and Shims. This is much easier to do with 35 degrees from best to worst, even considering the 2 degree resolution on Nvidia Mempry Junction Temperaturws within HWINFO.
@@snarksdomain Thanks for the answer :) I bought honeywell from aliexpress and moddiy a few months ago and it came with different foils, PTM from aliexpress came with grey foil and PTM from moddiy came with blue foil, their consistency was also different, PTM from aliexpress was easier to put on, to put PTM from moddiy I had to put it in fridge for 30 minutes first :D I came to the conclusion that all PTM from aliexpress are fake, they feel more like Upsiren than original Honeywell but they perform almost identical at least for the first few months, what is your opinion about this? Do you also have PTM from aliexpress to compare? Regards :)
@@snarksdomain Thanks for the answer :) I bought honeywell from aliexpress and moddiy a few months ago and it came with different foils, PTM from aliexpress came with grey foil and PTM from moddiy came with blue foil, their consistency was also different, PTM from aliexpress was easier to put on, to put PTM from moddiy I had to put it in fridge for 30 minutes first :D I came to the conclusion that all PTM from aliexpress are fake, they feel more like Upsiren than original Honeywell but they perform almost identical at least for the first few months, what is your opinion about this? Do you also have PTM from aliexpress to compare? Regards :)
Checking Your latest google drive charts it looks like PTM7950 has better temps now a days vs PCM-1. Right?
It seems that way at first glance of the charts, and it's possible that PTM7950 outperforms PCM-1.
With my current method of testing I believe I am out +/- 0.5 degrees Ceclsius, or around that. You will see results jump around between the CPU tests and the GPU Core tests. I'm hoping to make my testing more accurate in the future. There isn't enough of a spread with Pastes and Phase Change Materials results to be confident on the findings for me.
With Thermal Pads and Putties there is a much larger difference in results so it makes it easier to see a difference in performance. Even still, pad/putty results jumo around between ETCHASH, Furmark, and KAWPOW results.
I'm starting to form how I want to set up Testbench 3.0 in my mind. Will likely wait for RTX 50 series so we can test GDDR7 and see how hot it gets.
@@snarksdomain your test are good enough:) really great work. After checking Your charts I decided to go for CX H1300 13.5W instead of upsiren u6 pro, and ptm7950 instead of pcm-1 based on Your latest charts. I can use ptm for gpu and cpu if I am not mistaken right? Cheers
@@snarksdomain You do a great job. Based on Your results I will pick PTM7950 over PCM-1, and CX-H1300 over UPSIREN U6 PRO. That make sense do You think? BTW I can use PTM7950 for both CPU and GPU right? cheers
@smokehead5318 For sure. You can definitely use PTM7950 on both. I recommend giving it a good burn in for 2 hours. If you can get CPU temp up to around 65-70 Celsius if should get a good melt and become nice and thin. I saw the largest improvements in the first 2 hours for most of the PCM's, the exception being Laird TPCM7250, which needed a hotter burn in to get a proper melt (I'd recommend 85-95 Celsius burn in for Laird).
@@snarksdomain ok thank You for a such informative reply. Cx-h1300 is superior to u6 pro based on Your test, however it’s a bit stiffer during application?
Any plan to try out upsirentg-100 lol wonder how good are rhey
For sure. Once I get through all these thermal putties I'll be testing out Upsiten LMTG-100 and then Thermal Grizzly Conductonaut.
After those I'll do some thermal pastes (all the ones I have on hand), and then the Phase Change Materials. PCM's will take the longest to test.
@@snarksdomain thankyou sire! They claim there wont be "leak" like your typical lm (or thats what i read lol)..... Also speaking of pcm.... Liano also made a pcm that look identical to upsiren(the ads rofl) but appereanly it has "up to 13.8w/mk" might be worth a try lol
best thermal paste for laptop cpu and gpu if u can recommend
Phase Change Material. Either Honeywell PTM7950, Upsiren PCM-1, Thermalright Heilos, or even Laird TPCM7200 (requires higher temperature to melt than the other 3)
@@snarksdomain but i was asking only paste and not ptm, since in India there is issue to get anything form AliExpress its banned
@@snarksdomain can u recommend any paste and not ptm. tell me about some good paste, since in india lots of aliexpress does not deliver
Freeze the pcm material before appling to the cpu.
@edwardscroop9149 you can. These days I just use it a room temp, and cut it smaller by 3mm on all 4 sides. Makes application super easy, and 2nd peel no issue.
what's the benefit?
@@xCheatah Freezing might make it easy to cut with scissors without squishing it. But I haven't bothered freezing PCM's for a while.
Upsiren LMTG-100 is out can you try it please?😊
A sample of it is headed to me
hey! i just got 5900x. how is your 5900x ccd1 matching ccd2 temps under full load? mine doesnt... it differs as much as 20c temp...
I'm using Ryzen Master to set a fixed all core OC (frequency and voltage). Perhaps only one of your CCD's is working during the load you are using.
@@snarksdomain Can you please retest Copper shims? You only have tested them in the obsoelete chart, i want to see whether it's best to keep the current 3d graphite pad on my smartphone soc or replace with copper shim and paste?? Your videos are actually helping alot, thank you for your work.
@hotsauce6310 I'm testing out Aluminum Nitride shins and Putty roght now and will test the other shim and putty combos soon.
One nice thing about the AIN shims is that they don't conduct electricity.
@@snarksdomain I'm not really sure where to find those. I did not find graphene pads to cause any harm yet, i guess its the little project of mine, seeing it's longevity and it's reliability on my older phone. Seems to perform better than i remember and compared to others. Not sure by how much, note that i had to build the phone from scratch after 4 years of no use, maybe you could drop some insight on it? what about your obsoelete chart? There the graphite pads are no better than thermal putty. Wrong methodology used? or Pads size? I've seen u have to use higher than actual thermal pad size because graphite pads performs better when compressed. In my case i took off some top shielding that m/b had over soc & some other components and did a direct IC(Ram stacked on top of Cpu) to frame/housing application.
@@snarksdomain I saw somewhere that graphite pads only last around 6 months? is that true? What happens to them after it they get dry or does thermal conductivity drop significantly?
Still waiting for Upsiren lmtg 100 that h
Hi guys! I want to make a really low power setup out of a ryzen 4750ge I got my hands on a few weeks ago. Would this get hot enough to get the ptm to melt?
You might be able to get it hot enough if you play around with the fan speed and run something like Prime95 for a while.
@@snarksdomain thanks! I really appreciate the content you do, amazing stuff. Thank you!
Keep the pump off for the burn in
Can you tell if is enough the 40x80mm size for the cpu and gpu?
40x80 is definitely enough. If intel CPU, you'll likely be using 30x40mm. That would leave you 50x40mm, which is lots. You might even b3 able to mess up the CPU application once and still have enough for GPU.
If AMD then you'll be using closer to 40x40mm and will still have lots for GPU, but try your best to get the CPU application correct the first time. Likely best to do CPU first and if it messes up you might be able to salvage a big enough piece for GPU. Even if you have to use a few broken pieces, I would still go through with it. This stuff will melt and make a nice even layer in the end, as long as you have enough to melt and spread.
@@snarksdomain thank your for all the explanation and tips 🙏🏼
Way more than enough. You don't need nearly as much PTM7950 as you think. They come very thick and it melts to a super-thin layer from 200 microns down to around 20-25 microns, so all the extra thickness (volume) that you will have will be simply squeezed out. Best to put like a 20x20mm patch on a 35x35mm IHS, or a 17x17mm patch on a 20x20mm GPU die, because they heat up a lot harder and you shouldn't try to save up too much money on applying PTM7950. If it's a 35x42mm IHS like with Intel LGA1700 processors, just use a 20x24mm patch.
@@one_step_sideways THis is a good point. I'used large squares simply because the smaller sheets I got aren't large enough to have enough excess for other application.
I do have 1 large sheet of PTM7950 that I bought off Aliexpress and need to test to see if it's is real, but it looks real from the packaging and labelling. I'll definitely be cutting smaller pieces from sheet.
@@snarksdomain I wonder if you can put something like 4 small patches that are like 8x8mm and place them 5-10mm apart to guarantee the best spread without using too much paste by simulating an X-pattern like with thermal paste, since the spreader method is expensive. Because with a 20x20mm patch in the middle you have the same kind of spread as with a drop of paste, and it may still be too much, considering the 5-10x reduction in thickness. I think you should test that with your sheet of cheap PTM7950 you got.
You can put the 8x8mm patches on the IHS like so:
____________
| |
| × × |
| |
| × × |
|_________|
Tôi thực sự là fan hâm mộ của bạn. Tôi cũng đang điều hành một doanh nghiệp dịch vụ vệ sinh máy tính xách tay tại Việt Nam. Tôi có thể liên hệ với bạn ở đâu để hỏi một số câu hỏi liên quan đến mod kim loại lỏng và đồng vram
You can contact me at snarksdomain@gmail.com for questions if you like. I don't have experience with Liquid Metal yet, so will be studying the subject more before I go to test it.
Using phase change material on a CPU is a waste, even for testing purposes. Despite CPU using about 150W of power, this does not translate into 150W of radiated heat. All new Ryzen CPUs, be it 7nm chip models or newer, have very small die size, which limits heat output. My personal guess, about 80W gets out - that´s why with these CPUs, low-end 120mm tower coolers are able to compete with high-end 140mm dual towers. Even intel is on the same boat now, where latest gen chip is so dense, despite consuming lot of power and running hot, can´t be cooled down with anything.
Also, even small changes in ambient affect the results.
I´d steer away from the Upsiren too. It´s an unknown brand with unknown history. The product claims to be related to "computer-systems gr" printed on the box, which is actually k5 Pro putty manufacturer and they themselves distanced from Upsiren product brand. Even when it seems to work just fine, there is no documentation for... anything. Chinese sellers also took your data and are now claiming "our products have been tested", aka they themselves had no idea, how good/bad the stuff actually is.
I think it's worth testing Phase Change Materials, at least I'm having fun doing so.
I'll be testing more accurately with ambient temperature being logged using an Aquacomputer Quadro.
Upsiren is a good brand from all my testing so far. I agree, they are not connected to Computer-Systems.gr, thankfully. I'm not entirely sure why Upsiren listed that website on their packaging. I assume it was an attempt to help break into a competitive market of Thermal Interface Materials. That being said, I think they can stand on their own two legs now, and should likely remove said website from their packaging, to make things less muddy.
It's good to remember that most Thermal Interface Materials are manufactured in countries such as China, Malaysia, Vietnam, and South Korea. It is not unusual to have new companies start to create new materials for an ever growing market.
From talks with Upsiren, they have done their own internal testing of the products they manufacture. and there are some new products coming that I will be testing out as well.
That being said, I'm also looking forward to testing out more products from Fujipoly, Laird, Mg Chemicals, Thermal Grizzly, and more.
@@snarksdomain it´s usually a tactic to confuse customers into thinking, they´re buying stuff made by renown manufacturer. A good quality product can speak volumes by itself and does not need aggressive (fake) advertising.
It´s not surprising, everything is manufactured in East Asia. What is concering, however, is the total lack of documentation. What are the properties? Temperature for use? Temperature for storage purpose? etc.
@Morpheus-pt3wq Yeah, it would be better if Upsiren removes the Computer Systems website from the packaging.
Other than the basic specs, it would be nice to see some more documentation for sure.
I'll be doing some longevity testing of a several different products but won't have those results for at least a year (perhaps I should do confirmation testing in 1 year intervals and see how the products hold up.).
@Morpheus-pt3wq All of the heat the CPU generates must get out. 150W in = 150W out. It just takes a larger temperature difference between the die and cooler when the die is smaller. Combine that with the insane power consumption of the highest end models and you have trouble cooling them.
These testing vids are a great. Once the data logging gets improved, even better. No one else compared so many brands of putty and pads.
that's not how that works. that's not how any of that works
You got discord?
I do, but I don't share my Discord publicly at this time. You can email me if you like. Snarksdomain@gmail.com
𝙉𝙞𝙘𝙚. 𝙄'𝙢 𝙗𝙪𝙮𝙞𝙣𝙜 𝙩𝙝𝙞𝙨. 𝘾𝙖𝙣 𝙮𝙤𝙪 𝙘𝙤𝙢𝙥𝙖𝙧𝙚 𝙩𝙝𝙞𝙨 𝙩𝙤 𝙇𝙇𝘼𝙉𝙊 𝙥𝙝𝙖𝙨𝙚 𝙘𝙝𝙖𝙣𝙜𝙞𝙣𝙜 𝙩𝙝𝙚𝙧𝙢𝙖𝙡 𝙥𝙖𝙙 𝙖𝙨 𝙬𝙚𝙡𝙡? 𝙒𝙤𝙪𝙡𝙙 𝙡𝙤𝙫𝙚 𝙩𝙤 𝙨𝙚𝙚 𝙞𝙨 𝙩𝙝𝙚𝙧𝙚 𝙖𝙣𝙮 𝙗𝙚𝙩𝙩𝙚𝙧 𝙤𝙥𝙩𝙞𝙤𝙣𝙨 𝙩𝙝𝙖𝙣 𝙩𝙝𝙞𝙨. 𝙏𝙝𝙖𝙣𝙠 𝙮𝙤𝙪.
For some reason your comment was stuck in "Held for Review", which is not something I often check. My appologies for not responding until now. I will look up that LLANO Phase Change and see if I can get some for testing. Thanks for the suggestion.