@LockheedMartinF-22Raptor-y8i UTP-8 is better than UX Pro in every way. Performance/price/softness. If the laptop has very thin gaps, then I'd use something like Halnziye HY236. Most up-to-date charts are located at the Google Drive link in video description (Test Chart Repository > VRAM TIM > ETCHASH)
Thanks for all your work man! Doing us a great service with your comparisons. After using TG-PP10 for the last couple years with good results, It's awesome to see that there are cheaper options that perform even better. Ordered some of the U6 Pro, based on your data. I'm definitely curious to see if the gap between U6 and UX shrinks once you re-test with clean filters. Cheers!
Saw the LTT video on K5 pro and was intrigued. Upon further investigation, the reviews seemed quite mixed so I wasn't sure which thermal putty to buy. These tests you're doing make everything so much easier to understand. I'll probably be looking to get UX pro. you've earned yourself a sub
Thanks for checking out my videos and for the subscription. UX Pro is a great putty for Videocards. It's important to apply some force manually during reassembly to ensure it gets fully compressed. For any applications that have low mounting force, it might be worthwhile going with a softer putty, although I've used UX Pro on my laptop without issues. Upsiren will likely be making a softer version of this putty in the future.
That's some valuable content on your channel, I'm glad I found it! As you got no temperature controlled environment, it would be good if you added the delta from room temperature to the measured temperature for comparison.
Excellent point. I've since added ambient temperature sensors using an AquaComputer Quaddro. These temperatures show up in HWINFO64, which I use to record a log of all the computers sensor data during the tests. I then use Excel to calculate the delta values and use those results in the charts located at the Google Drive link below (Test Chart Repository > VRAM TIM > ETCHASH).
Another interesting thing to compare imo, would be silicone oil bleed. When I tried K5 Pro for my mobo VRM a few years ago, it bled so much silicone oil that I had streaks running down the entire length of the mobo. It got underneath the PCIe slots and was a huge pain to clean up. With TG-PP10, I noticed some bleed but it didn't seem worse than typical silicone thermal pads. Not sure how you'd do such a comparison. Maybe equal volumes of putty on some kind of absorbent paper, heating / baking the samples for a while, then measure the size of the oil spot on the paper.
I may try doing a bake test for this. Maybe I could try 212 Fahrenheit (100 Celsius) on the oven and see if anything happens. I believe K5 Pro at least would have a reaction at that temperature. then I could gradually bring the temperature up to 150 Celsius and probably stop there. I don't want to get too close to 200 Celsius.
@@snarksdomain Sounds like a solid plan m8! Yeah I think keeping it around the temp range seen in normal use, would be good. 150c sounds reasonable, if you're trying to accelerate the aging / oil bleed.
Just wanted to say thank you for all the work you put into making these videos. I went down a rabbit hole recently trying to figure out what to replace the paste on my laptop's vrms with and I really wanted to avoid K5 thermal compound. I was lucky to find some TG-PP10 after finding out that it was discontinued but was worried that there wasn't anything affordable to replace it. I look forward to every video you put out and thank you for sharing your findings!
I try not to take Thermal Conductivity ratings at face value anymore. If you watch one of my older videos, I test of TG-NSP80 (Electrically conductive so not safe). I test it out on the hotplate and it outperformed most other hotplate test candidates. It is apparently rated at 8 W/mk, but easily beat other pads and putties that were rated much higher. From my current view, w/mk ratings are marketing only, and something that all manufacturers are alomost fortced to play with, since most potential customers look at the one number to decide what to spend their money on. It's equivalent to a pizza resaurant saying "Best pizza in town"
@@snarksdomainMakes sense. Saw a thermal pad the other day that had 18w/mk and thought to myself that if it was real, why isn't everyone using it or as well known.
SuBBed. I was just looking into this and stumble on your video. Going to have to go back to learn understand how it's done and learn something. I've been eyeing out copper shims so will watch your other videos. Peace from South Africa 🇿🇦
I should be able to start retesting in the next couple weeks. Will probably do a short video showing the new setup vs old setup. I think the new setup is going to be a lot more reliable and consistent.
@@snarksdomain re. the ambient temperature issue; a lot of reviews for cooling stuff report the temperature difference between the ambient temp and whatever is being cooled rather than the absolute temp of the device. That takes the ambient temperature out of the equasion.
@spankeyfish I'll definitely be doing that going forward now that I have an Aquacomputer Quadro and it can be logged in HWINFO64. Looking forward to updating all my test results with better accuracy in the near future.
Hey! I've watched your videos and given this a go after the k5 pro turned into cheese on the higher temp parts of the card. So far everything it's supposed to be :) I hope it lasts longer than that the k5 XD Just a tip if you want it to stick a bit better while applying, rubbing it on the parts leaves that powdery residue you were talking about, it sticks to the residue a lot better.
zdravim super video ! mužu tě poprosit o radu ? mam notebok asus tuf gaming a15 fa506iu ? Co mi doporučuješ na VRM, SMD, VRAM ? Mužu tam dát UPSIREN U6 PRO? nebo co je nejlepší volba ? Thermal pad 0,5 je asi velký ? Děkuji zaradu
You will want to take a look at the latest charts, following the Google Drive link below (Test Chart Repository > VRAM TIM > ETCHASH). Currently, I'd probably go with either Upsiren UTP-8 or Fehonda LTP81. For the CPU and GPU core I'd use a phase change material like Honeywell PTM7950, Upsiren PCM-1, or Thermalright Heilos. You can also watch the video called "Thermal Putty Guide" to see how to apply it.
Hey Snarks! You are my savior! I heard about thermal putty some month ago, and as i just bought a 7800XT, i wanted to try, and while trying to get information on putty, i found your channel. A big thanks for all the test you do. I will go for Upsiren UX pro with custom copper shims. (and some other modification, like deshroud, new fans, and active backplate.) You always use MX-4 paste on the die? Did you really test it against other paste? How much quantity of putty you recommend to buy to make a full GPU (Vram, VRM, backplate, etc...)? 20g? 50g? Thanks for your work and your answers
Hey Adrien! Welcome to #TeamPutty! Glad you are going to try it out. There a lots of great thermal putties out there. Here is a link to the current charts: drive.google.com/drive/u/0/mobile/folders/1CAJP7ojt3jaGxBrJNTzd1Kp5-gikdh_8?usp=sharing Have a look at the charts and consider which will fit your budget. The top performing putties are all somewhat clay like, and can be used with shims, but you will need to apply pressure to make it squeeze down to the thim layer required to not have core contact issues. With shims it may be preferable to use a low viscosity putty. Either way, lots of options to choose from. As for Thermal Paste, I only use MX-4 for testing purposes and to keep everything consistent. It is cheap enough for me to buy large quantities while I test out putties/pads/shims. I recommend you get some Phase Change Material (Sheet). You can either go with Honeywell PTM7950, or Upsiren PCM-1. I still need to test out Thermalright Helios, but I'm assuming it will probably have similar performance as the othet 2. I would recommend buying 50grams so that you for sure have enough. a 3090 will often use closer to 70+ grames with it's 24 VRAm modules. Since you plan to use shims you should be fine with 50 grams and still have some leftover (NVME cooling perhaps?). Lastly, if using shims you might want to either buy some conformal coating like MG Chemicals 422C or Kaption Tape (1MIL / 0.001") to protect around the VRAM in the unlikely event that the shims ever move on you (maybe if you drop a videocard onto the ground?)
Excellent vid, very exciting results! You mentioned that this has a similar consistency to TH949-1 - do you think there would be any issue using the UX Pro along with copper shims?
Yes it will. You may want to check out the current charts located here: tinyurl.com/4btsmwu6 If you decide to go with UX Pro, be sure to apply a bit of pressure when you reassemble. UX Pro is a little stiffer than some other putties. I've managhed to work with it fine, but it does take a bit more attention. If you go with somethiung like Upsiren UTP-8, CX-H1300, Upsiren U6 Pro, or Kalhziye HY236/HY256 then it will be softed and easier to work with.
Judging from the packaging, this UX Pro is made by the same company that makes K5 Pro, which is absolutely terrible. And yeah, definitely get a thermometer to test ambient temp as well.
Computer-Systems has mentioned in an old Reddit that they have nothing to do with Upsiren. Upsiren uses their domain to give people confidence when buying the product, although I don't understand why they do it if the product in question is too good, even better than the K5-Pro.
@Dark Anubis That makes more sense now. My employer does a similar thing in the work we do (energy sector) so we can access more countries to work in. Countries and customers that would otherwise consider us too small, despite having a top of the line product. So from that parallel it clicks for me.
@@snarksdomain Exactly, it all comes down to Marketing. Although in this case Marketing is about a good product and not exactly copies and garbage as usually happens. I am currently using the U6 Pro (I didn't know UX Pro came out), now that I see this video I just ordered. The U6 Pro has performed too well on my Laptop (MSI Raider GE76 12UH), although I don't use it for the CPU because I use Liquid Metal for the GPU and CPU instead. P.S: New suscriber
@Dark Anubis thanks for the subscription! Yeah, I don't recommend using a putty on the core/die. I am experimenting with it just to see how it does, but mostly "for science!"
Hi i just need your advice please. So im switching the paste on my GPU and CPU Asus G16 to Thermal Grizzly PTM. Now for the VRAM and the other components can i use the Arctic MX6 paste or should i be using the thermal putty? and then of course if its thermal putty which should i be using?
@yousseffilali4635 you shouldn't use a paste. Putty is designed to be a gap filler, whereas paste is only meant to fill microscopic imperfections. Putty is also cheaper per gram than paste most of the time. Whichever putty you get, you will want to buy 20g. You might use 15g. Currently the top picks for Putty are Honeywell HT10000, Upsiren UTP-8, and Fehonda LTP81. Of your laptop has gaps smaller than 0.5mm then I'd recommend Halnziye HY236 as it can compress thinner than the others. If you want to see how they all compare with one another than the charts are located at the Google Drive link in video description (Test Chart Repository > VRAM TIM > ETCHASH)
@@snarksdomain Thank you so much! I have searched for all the pastes you recommened but couldn't find them here. I found the UPSIREN Ultra 16.8w/mk Thermal Putty and UPSIREN 12.8w/mk Thermal Putty. Would any of those two work?
@yousseffilali4635 Your best bet is to search on Aliexpress. You'll likely find the best price per gram there. As an example, you can often buy 100g for ~$40 CAD. 20g should be around $10 CAD. For Honeywell HT10000 you would have to buy it from Eugenery-KH. You can also buy Fehonda putty directly from them by contacting them through their website, same with Halnziye. Be careful when looking at Upairen UX Pro and Upairen U6 Pro. There are 2 versions, one from Upsiren Chine (Aliexpress), and one from Computer Systems GR Greece (Amazon). There's an entire trademark war happening between the two companies. Feel free to watch one of my other videos explaining what's going on. ua-cam.com/video/xND1726G624/v-deo.html
Yes, you should be able to reuse putty. I think the real goal, though, is to get to the point where we don't need to keep opening the card. And I think we finally have that option with Phase Change Materials for the core. We are no longer stuck with Thermal Pastes for the core. So far, we have Honeywell PTM7950, and PTM7958, Upsiren PCM-1 (currently testing), and even Thermalright is coming out with one now (or could be a rebrand, hard to say).
@@snarksdomain thank you for your input and yes I totally agree with the goal, waiting on a phase changing thermal paste/pad. Also do you use mining software for testing thermals ?
@nerijus01 Yeah, most of my testing has been with minung software so far, specifically with a memory intensive algorithm (Ethash/ETChash). It gets the memory hotter than Furmark or Timespy can. I might also consider using the Kawpow algorithm for higher performing solutions to test out the top end of core and memory thermals at the same time.
Anyone know if TG-N909 or TG-S808 any good? If they even a putty, A7000 only one I can find that is for sure. Reason I ask about those two is 909 has been listed on a few forums but no one really said anything about it from what I can see. Though looking at viscosity probably just regular paste? Trying to work out what will work for me, though also don't know how many grams I'll need, wanna do two PS3 launch models, my PS4, PS5 and GPU. A general idea on amount of product used per device if only adding what's required for coverage would also be great to know.
TG-N909 and TG-S808 are thermal pastes, not putty. They might be able to be used on a CPU, but looking at the dataset I can't confirm if they are electrical insulators, so I would be cautious. A7000 is one of the many thermal putties that Digikey sells. I have yet to test it but will at some point when I can afford to buy some. Digikey carries Penchem products as well. TH949-1 is their top performing putty. I've tested it and it's great. I haven't used putty on consoles yet, but will likely do my XBOX 360 (only console I have). Which GPU do you have?
@Snarks Domain 3060ti, trying to get most out if it before I swap the main GPU on my system. Atm the GPU is biggest bottleneck in my system, kinda curious about the vram upgrade mods though not sure if I'd be able to change the bios to make that work or if it's really worth it over upgrading card beyond a little project down the line. And all good, ended up ordering TG-A7000 as it was only one I could get as most the sites are blocking access to my IP address without extra hoops, was considering product in this video or the EVGA putty but couldn't find any in the country and don't really wanna wait three weeks, can only guess these puttys don't sell well to consumers which does make some sense. Mean I didn't know the name of this stuff until yesterday, I was calling it thermal gel and could never find anything when I looked it up for last few years. Changed my search to one for overcloaking and down the rabbit hole I've gone 😅 I've never really like thermal pads much as it's costing me alot for projects to keep getting different sheet sizes
It's not designed for that, and I'm not recommending that people do it, but as an experiment, I've had TG-PP10 on the core of a 3080 for over a year. Temps aren't terrible. I may test it out with the higher performing putties we have now. My gut tells me we're better off using a phase change material like PTM7950 or PCM-1 or others.
It isn't recommended to do that. Putties have a larger minimum bondline thickness than a paste would. You would likely be better using a phase change material such as Honeywell PTM7950 sheet or Upsiren PCM-1 That all being said, I've used putty on the core of a GPU for over a year, and the temps were just slightly worse off. I'll be testing it out at some point again on this testbench (but not until a bunch of other testing). I actually disassembled that card on stream. It was an EVGA 3080 XC3 with copper shims. I had used TG-PP10 on the core, and it was still soft and pliable. About 50 minutes in is where it starts: ua-cam.com/users/livejHhbDrY6o0E?si=VSQMcnItmKc_U-6_
@nguyenanghoang7314 the Upsiren thermal pads? I think they could be good. I had some of the 21w/mk ones on the testcard just trying them out and they seemed to perform well, but will need to fully test them once I go to test thermal pads. They don't really compete against PCM-1 as they are two different solutions. One as a gap filler. And one to melt down to a super thin bondline for CPU-Cooler heat transfer.
I doubt it reaches the performance of honeywell but it could definitely be very close considering how good upsiren's putty has been! It would actually be really interesting to see this!
@@treetop1496 From their ali description. Its exactly the same as the honeywell on max temp but the pcm-1 had 10C higher lowest temp. The phase change in the pcm-1 happens 10C higher than in the ptm7950.
hi, its good if i use it for stealth 15A13VF in 2023, how much quantity of putty you recommend to buy ? and what i can use thermal silicon for cpu and gpu ? my system i7 13620h 4060 8GB thanks for your help
I'm currently testing out some lower viscosity putties from Upsiren. This UX Pro can work in a laptop, but special attention is needed to make sure full compression is achieved. It may be easier to use a softer putty for laptops. As for the GPU core/die and CPU, I currently recommend using a Phase Change Material such as PTM7950 or PCM-1 (both are sheets), or you can also use PTM7958 (paste style), but it required extra steps of letting it cure for 24 hours after spreading a thin layer. Much harder to work with than the sheet type, in my experience.
@ThuanTran15219 I haven't tested that yet, but recommend looking into Honeywell PTM7950 or similar Phase Change Material (Upsiren PCM-1, Thermalright Heilos, etc.). Try not to get caught up on w/mk ratings. They are not often representative of perfoemanc3 and are mostly used a marketing.
I very much would love to see such a video where you put all the good pastes and thermal pads to the test to see once and for all what is better, putty or pads and which brand rules above the rest. Btw, have you ever tested arctic tp-3 pads?
I have a previous video which does a decent direct comparison between several putties pads and shim combos that might interest you. ua-cam.com/video/zSX_7P0HZqc/v-deo.html&t I will make another video at some point with more putties, pads, and shims included. For that video I will keep the paste the same for each test. I will need to make a separate video showing different paste/PCM options.
@@snarksdomain Would be a great help to many I am sure. Especially those of us with gaming laptops. Always a hassle to find the right size pads, if they even use pads anymore as quite a few manufacturers are stepping over to putty instead. Looking at the chart at the end of this video, I'd say the best option would be Upsiren ux pro + copper shims + mg chemicals 422C. (or just the ux pro if you want to avoid anything conductive.)
@@rick5078I've got updated charts located at the Google Drive link below (Test Chart Repository > VRAM TIM > ETCHASH). The charts get updated as I test more products.
Thanks for this ✌ Have you ever considered using one of those large diameter syringes, like the ones used to refill ink jet cartridges instead of a the piping bag?
I used syringes when I first discovered putty. I had got a stiffer batch (older) of TG-PP10 at the time and found the piping bag was easier to use then. I currently only have a piping bag for TG-PP10 with a couple hundred grams left in it. The only other putty I have enough of to use a bag would be Jeyi 8100 with around ~240 grams. I will likely go back to using large syringes for smaller quantities of putties. Rolling by hand can work really well with fully sized gloves. I find the little finger gloves aren't great for rolling into rope like strands.
I haven't tried preheating it. It compresses fairly easily, and I haven't had core contact issues (with the exception of that U6 Pro video where I used WAY too luch putty). The putty itself doesn't conpress like a pad does. It would be interesting to see if it flows easier or not when warmed, but I don't think it's designed to change viscosity when heated.
Just wanted to say that your channel is great. Honestly far better data than a lot of really popular channels. I'm looking to address some thermal issues on my XPS 7590 - or at least mitigate them a bit as it's getting quite old. I'm going with PTM7950 for the CPU & GPU but I'm also debating if I should change the stock thermal pads. Is putty tangibly better than after market thermal pads? Some years back I did a mod where you basically stack thermal pads to dissipate some heat from certain MOSFETs to the chassis. It did help but I wonder if putty would be more effective than stacking thermal pads to make up the height 🤔
Thanks for stopping by and your kind words. I think putty will worl well for you. I have an older chart comparing putty/pads/shims located in the VRAM chart section at the link in my channel about page. It's now obsolete since I wasn't tracking ambient back then, but it will give you a rough idea of relative performance between ourty and pads. I'll be retesting pads with my new setup when I can.
@@snarksdomain PTM7950 has now arrived along with some CX H1300 13.5W/m.K (seems liked a good compromise with cost vs performance for my budget). Do you have any suggestions or rules of thumb you use for amount to apply? I'll probably just be using some gloves and rolling it by hand. Edit: Just found your response on your most recent video with instructions. Will follow those, thanks again!
@gumbislice1605 that's what I though too, but I wasn't getting good enough compression. In the future I will opt for thinner shims to make it easier to compress. If you take a look at the most up to date charts you will see there isn't a corelation between mempry temperatures and core temperatures. This is because the each VRAM only produces around 3-4 watts. The charts are located at the Google Drive Link in the video description (Test Chart Repository>VRAM TIM>ETCHASH).
Your 3080ti should have 12GB of GDDR6X. If you only want to put putty on VRAM then 20g should be enough. On my test card I use around 14-15 grams on 8 VRAM (3070 ti). If you want to do putty on whole card and replace all thermal pads I recommend getting at least 50g to do the job. You will want to check the latest charts located on the Google Drive (Test Chart Repository). There is also a putty calculator there that you can use to enter your OEM pad dimensions and it will help calculate the amount of putty needed, and also the diameter of the logs to roll (if you use log method).
To my knowledge, UX Pro Ultra is the current top performing putty from Upsiren. That is the putty tested here. Is there another putty you've come across from Upsiren? (Other than U6 Pro and UX Pro)
@AzeroxDoomraider To be fair there isn't a "UX Pro" that doesn't have the word "Ultra". There is only a "U6" and a "UX" putty from Upsiren. It does make it a bit confusing that they both have the word "Pro" in their names.
Hi Mr. thermal paste and putty master. I bought hy236 putty 20 gr for laptop do you think it will be enough? (with old stock putty). unfortunately AliExpress doesn't send putty's to my country so i couldn't buy U6 Pro
@silvatis8061 You'll need around 50g for a 3080. I'd likely go with Upsiren UTP-8 or Fehonda LTP81 these days. I used 44 grams on a 3070. That had 8 VRAM chips. Your 3080 will either have 10 or 12 VRAM chips.
@shinotabi 50g for most GPU's. 75g for a 3090. 20g for laptop. Be sure to check the most recent charts at the Google Drive link in video description (Test Chart Repository > VRAM TIM > ETCHASH)
@shinotabi that the wrong UX Pro. Maybe consider going Upsiren UTP-8, Fehonda LTP81, or even Honeywell HT10000 from. EUGENEY_KH (youtuber). Even CX-H1300 13.5w would be a good option. You can buy Fehonda pitties directly from the ma factures via their website contact. Same woth Halnziye putties. Also Laird T-Putty 607 is another good option (Moddiy) Or see if this listing ships to Germany? www.ebuy7.com/index.php?route=product/product&product_id=479964582804248 You could even go with Halnziye HY236 or HY256 directly from Halnziye.
Hi, what are your thoughts on only replacing the vram pads on a 3080? Would it affect the temperatures if i only put the putty on the vram while the vrm and mosfets are the original pads? The reason i do not want to change them is because i read the 3080 cards uses other pads as spacers so having the right height is very important. Also my card is a very obscure card with not a lot of info regarding pads and in those sources people are claiming 1 mm pads and others 1.5 mm, so i just dont want to deal with that if i dont need to. Thank you in advance for all your work and efforts youre doing here, you deserve a lot of recognition for the quality of your content.
Thank you for your kind words. Which make/model of 3080 do you have. I don't see an issue in only swapping VRAM pads out. Generally those are the ones that bake/leak oil sooner anyways. And even if you decide to change everything out later, you can just reuse the putty by reforming it by hand, at least for a reasonable amount of time. I regularly only swap out VRAM interface materials during testing. In my case, I'm using TG-PP10 putty everywhere else (and MX-4 on the die). We won't know the long-term performance/reliability of UX Pro for quite some time, but I'm fairly confident it's going to be as good or better than most pads out there.
@@snarksdomain Thank you for your reply. I keep replying but my reply is not uploading for some reason. Anyways, I have a gigabyte turbo 3080. Its a blower style, so it gets quite hot and loud at times. I am hoping this putty along with MX-6 will fix that. Ill be updating my results! I hope this putty serves well for many years to come with no oil leakage. I'll be buying a depth gauge and using it before putting this putty to determine once and for all what thickness i would need in case i need to change the other pads.
@Tree Top Interesting that it's a blower style. This Putty should definitely allow the memory to run cooler and hopefully slow the fan down enough for you to hear the game you're playing. I usually save the stock pads by putting them on some parchment or wax paper, folded like a book, and then into a zip lock bag.
@@snarksdomain It really is an interesting card. It is extremely small for a 3080. Nvidea have blocked production of any geforce turbo card because of "cooling", but considering the rtx30 cards horrible cooling, the real reason is probably for the same reason they cut off SLI, they seem to be a good replacement for their overpriced workstation cards. I noticed in a video you said you live in canada as well, do you take donations through interac etransfer? i dont know how to use crypto
Com licença uma pergunta, com seu conhecimento de pastas térmicas qual pasta térmica você indica para reparo em celular, por que estou com um poco x3 pro problemático tenho o cooler DL05 mais não e suficiente, agradeço a atenção e prosperidade no seu trabalho
Olá Victor. I watched a bunch of videos and read posts about that phone. It seems like many people have thermal issues with it. How to fix those issues depends on how experienced you are with phone repair, and how much modification you want to make. It looks like there is a gap between the processor and the heat pipe, where Xiaomi used some thermal paste, but not enough. It doesn't get fully compressed so the "thermal bridge" provided by the paste is not good enough. You could try just cleaning it off and applying more paste to make a better "bridge" for the heat to get to the aluminium frame. If going this route I would probably recommend PTM7950, from the syringe. PTM7950 sheet won't work well in this application as the gap is wider than the sheet is. Another option would be to try just cleaning off the paste Xiaomi used and apply a putty there instead, like U6 Pro. I would go with a sticky putty on something like this I think. UX Pro is more clay like. If you wanted to go further and take the copper foil tape off, you could apply some thermal paste, or PTM7950, between the CPU and the foil, then replace the putty that is on the memory chip with more than Xiaomi used (maybe try Upsiren U6 Pro). You could even use PTM7950 between the CPU and foil. I would only do this if you have access to copper foil tape or if you can peel it back very carefully without tearing it. Then thermal paste/putty/ptm on top of foil again. I've never disassembled a phone that far yet myself, but maybe some day :)
@@snarksdomain obrigado sério muito obrigado mesmo eu ando procurando respostas sobre isso a dias já que é um assunto pouco comentado, enfim vou usar seu conselho muito obrigado mesmo pela atenção desejo muito prosperidade no seu trabalho💚
My pleasure. For your reference, here are some of the videos and posts I read/watched about your phone to get an understanding of the issue: forum.xda-developers.com/t/thermal-pad-copper-mod-for-poco-x3-pro.4479421/ ua-cam.com/video/1Tkmqn2fFMg/v-deo.html ua-cam.com/video/Ixl8tkhLLjk/v-deo.html ua-cam.com/video/tXKVTUt5VHE/v-deo.html imgur.com/a/wCq5IPK
I wouldn't pick UX Pro for a laptop anymore. I'd pick a softer putty. As for longevity of the different thermal putties, it will be a while before we know for most of them. TG-PP10 definitely set the bar high. I recently opened up my 3090 which had copper shims and TG-PP10 for 2.5 years. The putty was still soft. Here's hoping some of the other thermal putties can match that.
@mehmetoz1932 it's one of the putties I'd pick for a laptop based on it's softness. It's far too new of a product to know it's longevity, but it was designed to have better longevity than Upsiren U6 Pro, according to Upsiren, and there are folks who have had U6 Pro in their laptops for a little over a year now. Honeywell HT10000 seems to have a good track record of 2 years so far, according to folks buying it from Eugeney_KH (UA-camr). That's where I got mine from. I also consider Laird T-Putty 607 to have a good track record thus far, as it's seen a bunch if use in Europe/Germany. Halnziye is another well established manufacturer of putty and really competitive pricing as well. You can actually contact Haknziye directly using their website and purchase from them. Their HY236 and HY256 and an excellent putty for laptops as it is very soft. It'a performance is more than adequate for laptops. It wouldn't hesitate to use it. Penchem's TH930 has the closest resemblance to T-Globals TG-PP10. Pemchem is also a well established Thermal Putty manufacturer and has their products listed on Dogikey even (depending on which country you're shopping from). Fehonda has been making Thermal Interface Materials for quite a long time as well. It will be interesting to see how their LTP81 and LTP65 hold up over the coming years. I don't know what the best choice of a putty is for you. A lot depends on where you live, which can limit your options (depending on country). It also depends on your budget and your willingness to try something new. If you want top performing, worh a decent track record, from a large manufacturer and good softness then I think Honeywell HT10000 is the way to go. We're already trusting the longevity of PTM7950 from Honeywell, certainly they know how to make a good Thermal Putty.
@mehmetoz1932 I don't really have a way of answering that question. I have 3 laptops. A Clevo P170EM, an Aorus 15P YD, and a dell laptop for work. Your best bet is to search forums or discord channels for your specific laptop model and see what others are getting wirh Honeywell PTM7950 on the CPU/GPU core and putty everywhere else. The putty doesn't cool the CPU, but it does allow for more CPU/GPU core contact/pressure to give the potential of a very thin layer of Phase Change.
Hi Snarks, thank you for your testing, found you on reddit and you’ve been very helpful. Might need your advice with something, i want to change the thermal putty on my ps5 , a while ago you were recommending Th 949 1, what do you think about upsiren u6 pro or ux in this case ?
I can give you the results from my 3090, that I already have installed UX Pro on the VRAM+VRM, and a PTM7950 pad on the GPU die. Both the GPU core and hotspot only dropped around 5-6C (from 72.5C and 93.0C, respectively, to 67.8C and 89.5C). That's to be expected honestly since hotspot is still the hottest reading in the die only. VRAM on the other hand had a huge drop of ~20C. I use Stable Diffusion to stress test the VRAM, and UX Pro dropped it from 104C to 84C. But in any case, using thermal putty is what allowed these hotspot temperatures. I tried pads before, and temperatures always crept up because the pads were never at the right height. If you want to lower hotspot temperatures, I'd look into undervolting first. On this 3090 I'm running 800mV@1755Mhz and -500Mhz on the VRAM (for my use case, what I really need is VRAM amount, losing about 3% performance doesn't bother me). That, combined with larger fans I mounted, and a some small heatsinks on the card's backplate leaves my 3090 running on full load at 59.1C GPU Core, 75.3C Hotspot, and 78C VRAM. So all in all, UX Pro was very much worth it for my card, and I will very likely always use putty instead of pads from now on.
hi, great video! can i ask you for some advice I have an asus tuf gaming a15 fa506iu notebook? What do you recommend for VRM, SMD, VRAM? Can I put UPSIREN U6 PRO there? or what is the best choice? Thermal pad 0.5 is probably big? Thank you for the advice
You can definitely out the Thermal Putty of your choice anywhere there was a thermal pad. Lots of good ones to choose from. Have a look at the charts located on the Google Drive (Link in the description).
@@snarksdomain Can I ask you what would you recommend as the best choice for this laptop for VRM, SMD, VRAM? I also wrote to asus support and they didn't help me there! I'm already desperate! I would prefer to put there Thermal pad DLTECH 15w/mk - 90x50x0.5mm. But I can't find anywhere that 0.5 is not too big. thank you for the advice
@tomastinka4998 I'd either go with Upsiren UTP-8 or CX-H1300. If you pick CX-H1300 then you need to make sure the pictures in the product listing have a checkmark in the 13.5w/mk box and not the 8w/mk box. They are both called CX-H1300, and the descriptions of both often state 13.5w/mk. The only different is the check mark. As you can see on the chart, they do not perform the same.
For a 3090 I would get 70 grams if doing the entire card. You might be able to do it with 50 grams. Which brand of 3090 do you have? If it's EVGA you can leave the grey putty on the VRMs as it is also a very high performing putty and can easily be reshaped.
I got 50 grams myself. In my 3090 I used the almost entirety of it on the heatsink side of the card, but take note I went quite liberal on the amount of putty I used, I preferred to use a bit more and make sure I had a good coverage, than trying to ration the putty and risk leaving some area uncovered (for the backplate side, the Gelid Ultimate pads I tried initially made good contact, and I saw no point in fixing what was not broken). If you want to do the full card in putty, I'd say grab an extra 20g, or even jump straight to 100g, but again it depends on the amount you need for an application. For instance my 3090 needed 3mm thick pads for the mosfet's backside under the backplate, that's already a fair chunk of putty if I wanted to replace them.
@@joaoassuncao1 Yeah, a 3090 eats up a lot of putty. Might be a good use case for using 2 putties. UX Pro on VRAM and U6 Pro everywhere else. It would bring down the cost substantially.
I'm using the copper shim+tg-pp10 setup on my 3080 tuf and it's been a huge improvement. I have no current intention of redoing my 4070TI but I thoroughly enjoy watching all of your hard work brought to us in video so that for future reference, I'll know what route I want to go and maybe get a stockpile in advance of it. Thank you!!
Copper shim+TG-PP10 is still a great combo and should serve you well for quite some time. Did you add 2 copper shims between the intermediate VRAM heatsink and the cold plate? I put one shim in place between the heatsinks but still need to cut a narrow (~8mm wide) shim for the other side to maximize heat transfer "bandwidth". It seems like Nvidia might have inproved cooling of the VRAM for the 40 series. Ghost was mentioning his 40 series card seems to have good thermals as well. Hopefully you won't need to open your 4070Ti for a long time.
@Snarks Domain I bought a kit that had about 12 different thickness shims and each was almost a perfect fit over a vram chip. From chip to plate, it was 1 shim with the smallest amount of putty holding it in place, whichever shim took up the most space without flexing the heatsink. For the space between the 2 heatsinks, I had to stack appropriately sized shims to equal the space since I had ran out using them over the vram chips. Again, used putty to hold it in place. Usually that was less than a half grain of rice size. Even with calipers, this was all trial and error and I had to do this 2-3 times to get it all correct. Too thick and flexed the gpu out of contact too much causing overheat and shutdown or too thin and shifted out of place. Definitely some pucker factor thinking I bricked my card a few times.
@@snarksdomain i used the upsiren uxpro putty on the vram and mosfets the mosfets i made a thin log along them and the vram i made balls and squished them down abit to cover the full vram still with enough on to squish down furthermore with cooler pressure and for the die i used honeywell ptm 7950 I am using a sama im01 case its basically a nr200 knockoff with matx support i want to try the morpheus 8069 cooler by rajintek but there is no info for 7000 amd series compatibility
Ordered. Should receive it in a few days. 75cc should equal 258.75g which comes out at 1.78g/$1 CAD. The density being near that 3.5 mark might mean it's a good performer as well.
I found out yesterday my order of A7000 was canceled mid-shipment. So I ordered it again but now won't have it for a couple weeks. Let's us know your thoughts once you test it out. The more people we have testing putties, the better the data, and more reliable the information/recommendation to use putties instead of pads.
Hi. Again an excellent review. Now I want to order one. But how can I tell which one is the original in Ali Express. Can you help me with that? The link in the description section sends directly to the main page. Can you share an original product link? 🙏
When I click on the link it takes me right to the product. I'm not sure what the issue might be. You can also go to the MIddle999 Store and search within the store for "UX Pro". 2 results pop up, both of which are the correct product. One has a lower price, but higher shipping cost so that they work out to being nearly the same price overall.
@@snarksdomain Thanks. I ordered. I hope it reaches my hand quickly. I also have an Msi Ge 75 8se model laptop. When I removed the cooling blog to change the thermal paste, I saw that the paste on the vrrams had not changed at all since the date it was taken. I put the thermal pad of the floor cleaning them, closed the 1mm thick lid and when I worked, the temperature quickly hit 95 degrees. I understood that where there is putty, you need a pad where there is a putty pad. And I think I ordered the best paste on the market.
@@halitay I will probably make a video of changing the putty on my laptop. I have a Gigabyte Aorus 15p xd-73us224so Laptop. Currently it has TG-PP10 in it as I hadn't known of other putties when I first got it.
@Halit I will use this Putty on my laptop for the VRAM, VRM, etc. But will use normal thermal paste or perhaps Honeywell PTM7950 on the GPU Die and CPU Core.
friend, what would you recommend for a gaming laptop ( vram, gpu 2070 q-max) but with alliexpress? I bought a k5 pro but refrained from exchanging ... UPSIREN U6 PRO??
Hello Adrian. I take it your 2070 MAX-Q has a TGP of 80-90 watts? How are the thermal on it for you so far? I would recommend either using U6 Pro or UX Pro. I've currently got UX Pro on my gaming laptop (130w TGP 3080), and am using Honeywell PTM7950 sheet on CPU and GPU die. Temps are good so far and the PTM is amazing on the cores. So much better than the dried out Kryonaut Extreme I left on for 1 year. We'll see if the PTM and UX Pro are still gojng strong a year from now. If you end up getting a large enough improvement on thermals are you considering running a high TDP BIOS for your laptop? Which brand of Laptop do you have? The folks over on the Legion Discord channel have been modding their Lenovo laptops to good effect using various putties, shims, etc. It's a good place to check out what others are doing.
@@snarksdomain Wow, thanx for fast replay I have GX701GX woth i7 9750h I honestly don't know how to check the card's TDP (hmmm i think is above 115w) but the processor is 85W, I just ordered a honeywall and it's already there I wanted to replace everything, including thermopads, that's why I ordered the most recommended k5 pro :/, I will have to wait more than a month for the u6 pro, the u6 pro is currently in stock super promotion so I would choose it ($ 12 for 50 grams u6 pro is probably a great price)
I'd pick a different Putty than Upsiren UX Pro these days. There are cheaper and better options. Check out the charts located at the Google Drive link in video description (Test Chart Repository>VRAM TIM>ETCHASH) For Laptops: buy 20g For Videocards: Buy 50-100 grams. 50g will do most cards. Some cards like an RTX 3090 will need closer to 75g due tro VRAM on the backside.
Yeah, I had originally tested U6 Pro before UX Pro, but that video got taken down. I've since been retesting all of the putties. Feel free to review any of the live streams to see the application of each one, or follow the Google Drive link below to review the latest charts (Test Chart Repository > VRAM TIM > ETCHASH). Currently, the 2 best putties are Upsiren UTP-8 and Fehonda LTP81
It isn't the ideal product for GPU Die. For that I would recommend PTM7950 or another Phase Change Material, or normal paste. I'm not saying it won't work, but using a putty on a CPU or GPU core is highly experimental. I have used TG-PP10 putty on the core of a 3080, and it is still on it. I will perhaps test putty on core in the future, but there are many other videos I will need to make first. I'll be testing Upsiren PCM-1 at some point as well as Thermal Grizzly Kryosheet. testing Phase Change Materials will take time (likely 1 month per phase change material being tested).
Considering the price difference between U6 Pro and UX Pro, the UX Pro is not worth it for ordinary user. A question: How much putty is enough for 1 GPU?
If we compare UX Pro to the cost of pads it's actually fairly comparable and often cheaper since you don't have to buy 4 different thicknesses of pads, but I would agree that U6 Pro is a great bang for buck solution. Depending on the GPU 20g may be enough, but 50g would be a safer amount if replacing all pads on the card. If you just replace the VRAM pads, then 20g should be okay, unless it's a 3090 with its 24 VRAM modules.
@@snarksdomain just doing my 3060ti both front and back & everywhere (VRMs, coils, VRAM) consumed about 50g, with backplate consuming the most due to the gap there. As for how well it works, well... this GPU lacks any sensors, so i ended up putting some putty on my motherboard VRMs, which do have sensors.
@SebasTian-io1gw These days I'd pick other putties. Upsiren UTP-8 has similar/better performance, is softer and cheaper. If a laptop has gaps smaller than 0.5mm then something like Halnziye HY236 is a solid choice as it can compress down easier than UX Pro or UTP-8. The best performing putty I know of is Honeywell HT10000. It costs a bit more and not quite as accessible. Fehonda LTP81 recently came down in price a bit closer to UTP-8 pricing so that is also a good option. Updated charts are located at the Google Drive link below (Test Chart Repository>VRAM TIM>ETCHASH).
Thank you . I buy 50g of this putty , so you recomend the others puttyes because are much softer than this?. I dont have other alternative to use this on my laptop now :/. You recomend wait and buy the others puttys?
@SebasTian-io1gw it can work well if you've already bought it. I've only tested the Upsiren one so far, so hopefully you got that one, and not the CSGR version of UX Pro. I recently opened my laptop up after 16.5 months of having Upsiren UX Pro and it was still soft enough to mold by hand. Just slightly dryer than brand new, so the stuff definitely lasts fairly well so far. Up to you whether you stick with the one you already bought or if you can still return for refund. Upsiren UTP-8 is like half the price of UX Pro. Laptops only need 20g or less, so you can do a laptop for arpund $11-$15 CAD.
You likely won't be able to source Upsiren products within Europe for a little while. There's a trademark war going on..... ua-cam.com/video/xND1726G624/v-deo.html
@DragonXDrei it was definitely WAY too much putty. I've gotten better at using less. There's a Putty Calculator on the Google Drive link which can be used for calculating putty log diameter based on the stock pad dimensions.
Could you do a GPU test with K5 Pro? While it took a long time to reach the 100°C in the older test it would be interesting if that's the only variable that matters. The few times I've seen K5 Pro in real world uses it performed good, but it was never compared to other thermal putties. Would be very interesting to see.
As mush as I don't want to use K5 Pro on my test card, I probably should test it so we can have better data and have a real world comparison. I'll mostly do it to show why I think it's a bad idea, and save others the headache. While the 100 degree hotplate test isn't the final word, it has been fairly reliable in showing relative performance of different pads/putties and GPU tests usually line up in the same order of performance.
Don't have any experience with K5 Pro itself (well, other than my DAC/AMP, but that's from Fiio, not Computer-Systems), but from what I've seen around in other videos and forums in the web, the problem with K5 Pro is that it seems to have a relatively low melting point, with some people even reporting it gaining a bubbly consistency when it started to reach temperatures around 100C. There are some reports of good results, yes, but I wouldn't be surprised is those came from low power applications. I'd wager that K5 Pro's thermal conductivity drops like a brick once it reaches higher temperatures.
@AAisAA I'd pick a different putty than UX Pro. There are better/similar performing options for cheaper. According to online conce sus, for a 4090 you want to aim for ~60 Celcius Memory Junction Temeprature for max OC headroom. So keep that in mind when selecting a Thermal Putty. You'd want to see where your stock temps are and go from there. Aside from all that, yes, Thermal Putty is great for anywhere there was a thermal pad or other thermal putty. I recommend PTM7950, or similar Phase Change Material, for the core.
@@AAisAA pricing can be all over the place, depending on where you love a d cost of shipping/import. Check the chart located on the Google Drive link below (Test Chart Repository>VRAM TIM>ETCHASH)
I'd expect a guy dealing with such tedious testing on thermal interface materials to know the exact model of his computer case with certainty, because surely such a person would care about how cool his PC will run and that would require him to look for a case with adequate air flow, which means he'd have to make some research at least and then by the end of it he'd have the model etched onto his memory
Oh I do. Lol. My gaming rig is in a Lian Li O11D XL Black. Used to run my main PC in a Thermaltake Xaser III Super Tower (Blue). For a moment I couldn't remember if the Fractal Design was a Meshify 2 or a Meshify C.
@@snarksdomain I see. I have a suggestion for a video, why dont you pick any high performance thermal pad brand, and then use pads thicker than what your graphics card requires, and then compare the results to when you use the proper thickness thermal pads. The title of the video would be "Does thermal thickness really have to exact?" or something like that. I'm thinking that since thermal pads are generally very soft and squishable, the eventual thickness of a thermal pad once the graphics card is assembled would not depend on the original thickness of the thermal pads, as the pressure from all the screwing would most likely squish all the thermal pads to the 'right size' if any of them are thicker than they are 'supposed to be'. My graphics card is still under warranty and I'm (for now...) a pathetic coward who doesnt want to open up his GPU and 'risk' breaking something, so I cant make that test myself now. Also there is factor of it having taken me a traumatic amount of effort to fit the damn GPU in my case, so I ideally dont want to go through that again. (Case is Meshlicious, GPU is an MSI 3080 Gaming Z Trio)
@Andy96 surprisingly, that's basically how I ended up here. First with trying Gelid Ultimate pads and utterly failing and getting horrible temperatures and core contact issues. Then I bought a bunch of Gelid Extreme pads because they were supposed to be softer and more forgiving. I had mixed results even with the "correct" pad thicknesses everywhere. I think the main problem stems from 4-6 pad thicknesses on a card all needing to make contact and still get core contact for it to work out. The beat chance at making pads work out is likely to compress then by rolling them down to about 105% of the gap size. That way they only need that little bit extra squish from the mounting pressure for everything to make contact. If you like you can check out some of my earlier attempts at repadding cards with mixed results here www.reddit.com/r/nvidia/comments/pnwogn/asus_strix_3090_thermal_putty_tgpp10_instead_of/?
@@snarksdomain Then I take it that they arent 'that' malleable, because if they were you wouldnt have any problems when you used thicker thermal pads, they would have just comformed to the proper thickness under the mounting pressure. So somehow they have enough rigidity to resist the mounting pressure and bend the cooling plate, which in turn causes different parts of the PCB to lose proper contact with the cooling plate, causing high temps? This is theoretically very stupid and is very hard to believe. How tough could they possibly be? What could I be missing? EDIT: I'm also wondering, is the reported memory temp pulled from just one specific memory chip, or is it an average of the temps from all the memory chips? This could be tested by cooling every chip manually without the cooling plate...
@Andy96 that was my surprise as well when first trying to repad videocards. I thought that if I took accurate measurements or even followed other people guides wkth listed thicknesses then all would be fine. I bought a few hundred dollar of thermal pads (Gelid Ultimates and then Gelid Extremes). Extra thick pads won't bend the coldplaye or cooler. They are much too rigid for that. What will happen is you might not have any core contact at all. Since moving to putty I haven't had to worry too much.....except for when I used WAY too much putty during the Upsuren U6 Pro (About 3x as much as was needed). But there is a much larger margin of error when using putty. I will add that using pads on VRAM and outta everywhere else can work good too if you REALLY want to use a pad for VRAM. Just gotta be sure to get the correct thickness. But getting one right is easier than getting 4-6 right....especially when some EVGA cards use 2.25mm or 2.75mm pads. Memory temp shows the highest memory temperature.
Just bought this and was wondering, so glad to see your video and know for sure! Your the only source out there for good knowledge on thermal puttys
Thanks :) Hopefully you got it from Aliexpress.
@@snarksdomain😬Amazon; is it just a better deal from Ali or?
@@snarksdomain hi between UPSIREN UX PRO and UPSIREN UTP-8, which do you think is better for my vram and vrm laptop? asus tuf a15 ryzen 7735hs rtx4060
@LockheedMartinF-22Raptor-y8i UTP-8 is better than UX Pro in every way. Performance/price/softness.
If the laptop has very thin gaps, then I'd use something like Halnziye HY236.
Most up-to-date charts are located at the Google Drive link in video description (Test Chart Repository > VRAM TIM > ETCHASH)
Thanks for all your work man! Doing us a great service with your comparisons. After using TG-PP10 for the last couple years with good results, It's awesome to see that there are cheaper options that perform even better. Ordered some of the U6 Pro, based on your data. I'm definitely curious to see if the gap between U6 and UX shrinks once you re-test with clean filters. Cheers!
This channel is an unseen gem. Keep up the good work!
Saw the LTT video on K5 pro and was intrigued. Upon further investigation, the reviews seemed quite mixed so I wasn't sure which thermal putty to buy. These tests you're doing make everything so much easier to understand. I'll probably be looking to get UX pro. you've earned yourself a sub
Thanks for checking out my videos and for the subscription. UX Pro is a great putty for Videocards. It's important to apply some force manually during reassembly to ensure it gets fully compressed. For any applications that have low mounting force, it might be worthwhile going with a softer putty, although I've used UX Pro on my laptop without issues.
Upsiren will likely be making a softer version of this putty in the future.
That's some valuable content on your channel, I'm glad I found it!
As you got no temperature controlled environment, it would be good if you added the delta from room temperature to the measured temperature for comparison.
Excellent point. I've since added ambient temperature sensors using an AquaComputer Quaddro. These temperatures show up in HWINFO64, which I use to record a log of all the computers sensor data during the tests.
I then use Excel to calculate the delta values and use those results in the charts located at the Google Drive link below (Test Chart Repository > VRAM TIM > ETCHASH).
Another interesting thing to compare imo, would be silicone oil bleed.
When I tried K5 Pro for my mobo VRM a few years ago, it bled so much silicone oil that I had streaks running down the entire length of the mobo. It got underneath the PCIe slots and was a huge pain to clean up. With TG-PP10, I noticed some bleed but it didn't seem worse than typical silicone thermal pads.
Not sure how you'd do such a comparison. Maybe equal volumes of putty on some kind of absorbent paper, heating / baking the samples for a while, then measure the size of the oil spot on the paper.
I may try doing a bake test for this. Maybe I could try 212 Fahrenheit (100 Celsius) on the oven and see if anything happens. I believe K5 Pro at least would have a reaction at that temperature. then I could gradually bring the temperature up to 150 Celsius and probably stop there. I don't want to get too close to 200 Celsius.
@@snarksdomain Sounds like a solid plan m8! Yeah I think keeping it around the temp range seen in normal use, would be good. 150c sounds reasonable, if you're trying to accelerate the aging / oil bleed.
Just wanted to say thank you for all the work you put into making these videos. I went down a rabbit hole recently trying to figure out what to replace the paste on my laptop's vrms with and I really wanted to avoid K5 thermal compound. I was lucky to find some TG-PP10 after finding out that it was discontinued but was worried that there wasn't anything affordable to replace it. I look forward to every video you put out and thank you for sharing your findings!
Damn that 6.8W/mK difference between the UX Pro Ultra and TG-PP10 is doing a lot of work.
I try not to take Thermal Conductivity ratings at face value anymore. If you watch one of my older videos, I test of TG-NSP80 (Electrically conductive so not safe). I test it out on the hotplate and it outperformed most other hotplate test candidates. It is apparently rated at 8 W/mk, but easily beat other pads and putties that were rated much higher.
From my current view, w/mk ratings are marketing only, and something that all manufacturers are alomost fortced to play with, since most potential customers look at the one number to decide what to spend their money on.
It's equivalent to a pizza resaurant saying "Best pizza in town"
@@snarksdomainMakes sense. Saw a thermal pad the other day that had 18w/mk and thought to myself that if it was real, why isn't everyone using it or as well known.
Looking forward to seeing the next tests...
Ayy its actually better than u6 pro. Looks like i have a reason to upgrade mine. Thanks for the reviw kind sire
Check out Honeywell PTM7950 phase-change material by applying to the core.
Very useful tests you do thermal putty. Greetings.
Thanks for this vid man. I used this putty and tfx on my card and its 4-6degrees cooler now
thanks for your science.
I don't know why but those "logs" at 0:42 cracked me up.
Looking back at it now I also crack up at the thought of using that much putty. It was WAY too much.
SuBBed. I was just looking into this and stumble on your video. Going to have to go back to learn understand how it's done and learn something. I've been eyeing out copper shims so will watch your other videos. Peace from South Africa 🇿🇦
Awesome news. Evga putty been sold out for months.
looking forward to the open air test bench and retests
I should be able to start retesting in the next couple weeks. Will probably do a short video showing the new setup vs old setup. I think the new setup is going to be a lot more reliable and consistent.
@@snarksdomain re. the ambient temperature issue; a lot of reviews for cooling stuff report the temperature difference between the ambient temp and whatever is being cooled rather than the absolute temp of the device. That takes the ambient temperature out of the equasion.
@spankeyfish I'll definitely be doing that going forward now that I have an Aquacomputer Quadro and it can be logged in HWINFO64.
Looking forward to updating all my test results with better accuracy in the near future.
0:42 Next day chipotle size application.
Hey! I've watched your videos and given this a go after the k5 pro turned into cheese on the higher temp parts of the card. So far everything it's supposed to be :) I hope it lasts longer than that the k5 XD
Just a tip if you want it to stick a bit better while applying, rubbing it on the parts leaves that powdery residue you were talking about, it sticks to the residue a lot better.
Welcome to #TEAMPUTTY :)
What GPU are you using it on? Just wondering if you're able to see Memory Junction Temperatures or not.
zdravim super video ! mužu tě poprosit o radu ? mam notebok asus tuf gaming a15 fa506iu ? Co mi doporučuješ na VRM, SMD, VRAM ? Mužu tam dát UPSIREN U6 PRO? nebo co je nejlepší volba ? Thermal pad 0,5 je asi velký ? Děkuji zaradu
You will want to take a look at the latest charts, following the Google Drive link below (Test Chart Repository > VRAM TIM > ETCHASH).
Currently, I'd probably go with either Upsiren UTP-8 or Fehonda LTP81.
For the CPU and GPU core I'd use a phase change material like Honeywell PTM7950, Upsiren PCM-1, or Thermalright Heilos.
You can also watch the video called "Thermal Putty Guide" to see how to apply it.
Hey Snarks!
You are my savior!
I heard about thermal putty some month ago, and as i just bought a 7800XT, i wanted to try, and while trying to get information on putty, i found your channel.
A big thanks for all the test you do.
I will go for Upsiren UX pro with custom copper shims. (and some other modification, like deshroud, new fans, and active backplate.)
You always use MX-4 paste on the die? Did you really test it against other paste?
How much quantity of putty you recommend to buy to make a full GPU (Vram, VRM, backplate, etc...)? 20g? 50g?
Thanks for your work and your answers
Hey Adrien! Welcome to #TeamPutty! Glad you are going to try it out. There a lots of great thermal putties out there. Here is a link to the current charts: drive.google.com/drive/u/0/mobile/folders/1CAJP7ojt3jaGxBrJNTzd1Kp5-gikdh_8?usp=sharing
Have a look at the charts and consider which will fit your budget. The top performing putties are all somewhat clay like, and can be used with shims, but you will need to apply pressure to make it squeeze down to the thim layer required to not have core contact issues. With shims it may be preferable to use a low viscosity putty.
Either way, lots of options to choose from.
As for Thermal Paste, I only use MX-4 for testing purposes and to keep everything consistent. It is cheap enough for me to buy large quantities while I test out putties/pads/shims. I recommend you get some Phase Change Material (Sheet). You can either go with Honeywell PTM7950, or Upsiren PCM-1. I still need to test out Thermalright Helios, but I'm assuming it will probably have similar performance as the othet 2.
I would recommend buying 50grams so that you for sure have enough. a 3090 will often use closer to 70+ grames with it's 24 VRAm modules. Since you plan to use shims you should be fine with 50 grams and still have some leftover (NVME cooling perhaps?).
Lastly, if using shims you might want to either buy some conformal coating like MG Chemicals 422C or Kaption Tape (1MIL / 0.001") to protect around the VRAM in the unlikely event that the shims ever move on you (maybe if you drop a videocard onto the ground?)
Excellent vid, very exciting results! You mentioned that this has a similar consistency to TH949-1 - do you think there would be any issue using the UX Pro along with copper shims?
I think it will work great with shims. Will definitely test that in the future.
Upsiren is going to send me one of their copper plates to test it out with. Perhaps I'll test it side by side with my DIY copper and Aluminum shims.
@@snarksdomain Upsiren's becoming a one-stop shop for GPU cooling, I love it!
@@SouperRad They've definitely got momentum :)
Will this stuff work on a 4090 with a water block? I cant find the correct size pads for this anymore and the old ones are in bad shape
Yes it will. You may want to check out the current charts located here: tinyurl.com/4btsmwu6
If you decide to go with UX Pro, be sure to apply a bit of pressure when you reassemble. UX Pro is a little stiffer than some other putties. I've managhed to work with it fine, but it does take a bit more attention.
If you go with somethiung like Upsiren UTP-8, CX-H1300, Upsiren U6 Pro, or Kalhziye HY236/HY256 then it will be softed and easier to work with.
Judging from the packaging, this UX Pro is made by the same company that makes K5 Pro, which is absolutely terrible.
And yeah, definitely get a thermometer to test ambient temp as well.
That's interesting. I wonder if Computer Systems is a manufacturer or perhaps a redistributor.
Either way, UX Pro is a great putty.
Computer-Systems has mentioned in an old Reddit that they have nothing to do with Upsiren. Upsiren uses their domain to give people confidence when buying the product, although I don't understand why they do it if the product in question is too good, even better than the K5-Pro.
@Dark Anubis That makes more sense now. My employer does a similar thing in the work we do (energy sector) so we can access more countries to work in. Countries and customers that would otherwise consider us too small, despite having a top of the line product.
So from that parallel it clicks for me.
@@snarksdomain Exactly, it all comes down to Marketing. Although in this case Marketing is about a good product and not exactly copies and garbage as usually happens.
I am currently using the U6 Pro (I didn't know UX Pro came out), now that I see this video I just ordered. The U6 Pro has performed too well on my Laptop (MSI Raider GE76 12UH), although I don't use it for the CPU because I use Liquid Metal for the GPU and CPU instead.
P.S: New suscriber
@Dark Anubis thanks for the subscription!
Yeah, I don't recommend using a putty on the core/die. I am experimenting with it just to see how it does, but mostly "for science!"
Hi i just need your advice please. So im switching the paste on my GPU and CPU Asus G16 to Thermal Grizzly PTM. Now for the VRAM and the other components can i use the Arctic MX6 paste or should i be using the thermal putty? and then of course if its thermal putty which should i be using?
@yousseffilali4635 you shouldn't use a paste. Putty is designed to be a gap filler, whereas paste is only meant to fill microscopic imperfections. Putty is also cheaper per gram than paste most of the time.
Whichever putty you get, you will want to buy 20g. You might use 15g.
Currently the top picks for Putty are Honeywell HT10000, Upsiren UTP-8, and Fehonda LTP81.
Of your laptop has gaps smaller than 0.5mm then I'd recommend Halnziye HY236 as it can compress thinner than the others.
If you want to see how they all compare with one another than the charts are located at the Google Drive link in video description (Test Chart Repository > VRAM TIM > ETCHASH)
@@snarksdomain Thank you so much! I have searched for all the pastes you recommened but couldn't find them here. I found the UPSIREN Ultra 16.8w/mk Thermal Putty and UPSIREN 12.8w/mk Thermal Putty. Would any of those two work?
@yousseffilali4635 Your best bet is to search on Aliexpress. You'll likely find the best price per gram there. As an example, you can often buy 100g for ~$40 CAD. 20g should be around $10 CAD.
For Honeywell HT10000 you would have to buy it from Eugenery-KH.
You can also buy Fehonda putty directly from them by contacting them through their website, same with Halnziye.
Be careful when looking at Upairen UX Pro and Upairen U6 Pro. There are 2 versions, one from Upsiren Chine (Aliexpress), and one from Computer Systems GR Greece (Amazon). There's an entire trademark war happening between the two companies. Feel free to watch one of my other videos explaining what's going on.
ua-cam.com/video/xND1726G624/v-deo.html
Is Upsiren UX PRO Ultra thermal putty reusable ,say I open the card to change thermal paste can I then reuse that putty ?
Yes, you should be able to reuse putty.
I think the real goal, though, is to get to the point where we don't need to keep opening the card. And I think we finally have that option with Phase Change Materials for the core.
We are no longer stuck with Thermal Pastes for the core. So far, we have Honeywell PTM7950, and PTM7958, Upsiren PCM-1 (currently testing), and even Thermalright is coming out with one now (or could be a rebrand, hard to say).
@@snarksdomain thank you for your input and yes I totally agree with the goal, waiting on a phase changing thermal paste/pad.
Also do you use mining software for testing thermals ?
@nerijus01 Yeah, most of my testing has been with minung software so far, specifically with a memory intensive algorithm (Ethash/ETChash). It gets the memory hotter than Furmark or Timespy can. I might also consider using the Kawpow algorithm for higher performing solutions to test out the top end of core and memory thermals at the same time.
Anyone know if TG-N909 or TG-S808 any good? If they even a putty, A7000 only one I can find that is for sure. Reason I ask about those two is 909 has been listed on a few forums but no one really said anything about it from what I can see. Though looking at viscosity probably just regular paste?
Trying to work out what will work for me, though also don't know how many grams I'll need, wanna do two PS3 launch models, my PS4, PS5 and GPU.
A general idea on amount of product used per device if only adding what's required for coverage would also be great to know.
TG-N909 and TG-S808 are thermal pastes, not putty. They might be able to be used on a CPU, but looking at the dataset I can't confirm if they are electrical insulators, so I would be cautious.
A7000 is one of the many thermal putties that Digikey sells. I have yet to test it but will at some point when I can afford to buy some.
Digikey carries Penchem products as well. TH949-1 is their top performing putty. I've tested it and it's great.
I haven't used putty on consoles yet, but will likely do my XBOX 360 (only console I have). Which GPU do you have?
@Snarks Domain 3060ti, trying to get most out if it before I swap the main GPU on my system. Atm the GPU is biggest bottleneck in my system, kinda curious about the vram upgrade mods though not sure if I'd be able to change the bios to make that work or if it's really worth it over upgrading card beyond a little project down the line.
And all good, ended up ordering TG-A7000 as it was only one I could get as most the sites are blocking access to my IP address without extra hoops, was considering product in this video or the EVGA putty but couldn't find any in the country and don't really wanna wait three weeks, can only guess these puttys don't sell well to consumers which does make some sense. Mean I didn't know the name of this stuff until yesterday, I was calling it thermal gel and could never find anything when I looked it up for last few years. Changed my search to one for overcloaking and down the rabbit hole I've gone 😅 I've never really like thermal pads much as it's costing me alot for projects to keep getting different sheet sizes
@@Gikui right on. Let us know how A7000 does for you :)
I am seeing some aliexpress reviews of people using the thermal putty as thermal paste, I mean in the core itself, is this right?
It's not designed for that, and I'm not recommending that people do it, but as an experiment, I've had TG-PP10 on the core of a 3080 for over a year. Temps aren't terrible. I may test it out with the higher performing putties we have now.
My gut tells me we're better off using a phase change material like PTM7950 or PCM-1 or others.
Can I use a thin thermal putty layer to replace an old thermal paste on CPU and GPU core?
It isn't recommended to do that. Putties have a larger minimum bondline thickness than a paste would.
You would likely be better using a phase change material such as Honeywell PTM7950 sheet or Upsiren PCM-1
That all being said, I've used putty on the core of a GPU for over a year, and the temps were just slightly worse off. I'll be testing it out at some point again on this testbench (but not until a bunch of other testing).
I actually disassembled that card on stream. It was an EVGA 3080 XC3 with copper shims. I had used TG-PP10 on the core, and it was still soft and pliable.
About 50 minutes in is where it starts:
ua-cam.com/users/livejHhbDrY6o0E?si=VSQMcnItmKc_U-6_
@@snarksdomain thank you for your information. What do you think about Upsiren OEM? How they compares with Upsiren PCM-1?
@nguyenanghoang7314 the Upsiren thermal pads? I think they could be good. I had some of the 21w/mk ones on the testcard just trying them out and they seemed to perform well, but will need to fully test them once I go to test thermal pads.
They don't really compete against PCM-1 as they are two different solutions. One as a gap filler. And one to melt down to a super thin bondline for CPU-Cooler heat transfer.
Bro, can you test Upsiren PCM-1 ? It looks like Honeywell ptm 7950
I doubt it reaches the performance of honeywell but it could definitely be very close considering how good upsiren's putty has been! It would actually be really interesting to see this!
@@treetop1496 still waitting for this guy to test
@@treetop1496 From their ali description. Its exactly the same as the honeywell on max temp but the pcm-1 had 10C higher lowest temp. The phase change in the pcm-1 happens 10C higher than in the ptm7950.
hi, its good if i use it for stealth 15A13VF in 2023, how much quantity of putty you recommend to buy ?
and what i can use thermal silicon for cpu and gpu ?
my system i7 13620h
4060 8GB
thanks for your help
I'm currently testing out some lower viscosity putties from Upsiren. This UX Pro can work in a laptop, but special attention is needed to make sure full compression is achieved.
It may be easier to use a softer putty for laptops.
As for the GPU core/die and CPU, I currently recommend using a Phase Change Material such as PTM7950 or PCM-1 (both are sheets), or you can also use PTM7958 (paste style), but it required extra steps of letting it cure for 24 hours after spreading a thin layer. Much harder to work with than the sheet type, in my experience.
@@snarksdomain do you think about thermalright tf9 for cpu/gpu on laptop i checked it 14W/mk
@ThuanTran15219 I haven't tested that yet, but recommend looking into Honeywell PTM7950 or similar Phase Change Material (Upsiren PCM-1, Thermalright Heilos, etc.).
Try not to get caught up on w/mk ratings. They are not often representative of perfoemanc3 and are mostly used a marketing.
@@snarksdomain thank you so much
thanks, it was helpful and informative for me
I very much would love to see such a video where you put all the good pastes and thermal pads to the test to see once and for all what is better, putty or pads and which brand rules above the rest. Btw, have you ever tested arctic tp-3 pads?
I have a previous video which does a decent direct comparison between several putties pads and shim combos that might interest you.
ua-cam.com/video/zSX_7P0HZqc/v-deo.html&t
I will make another video at some point with more putties, pads, and shims included. For that video I will keep the paste the same for each test.
I will need to make a separate video showing different paste/PCM options.
@@snarksdomain Would be a great help to many I am sure. Especially those of us with gaming laptops. Always a hassle to find the right size pads, if they even use pads anymore as quite a few manufacturers are stepping over to putty instead. Looking at the chart at the end of this video, I'd say the best option would be Upsiren ux pro + copper shims + mg chemicals 422C. (or just the ux pro if you want to avoid anything conductive.)
@@rick5078I've got updated charts located at the Google Drive link below (Test Chart Repository > VRAM TIM > ETCHASH).
The charts get updated as I test more products.
@@snarksdomain sweet, on behalf of all those whom have a use for this, thank you very much!
Thanks for this ✌
Have you ever considered using one of those large diameter syringes, like the ones used to refill ink jet cartridges instead of a the piping bag?
I used syringes when I first discovered putty. I had got a stiffer batch (older) of TG-PP10 at the time and found the piping bag was easier to use then.
I currently only have a piping bag for TG-PP10 with a couple hundred grams left in it. The only other putty I have enough of to use a bag would be Jeyi 8100 with around ~240 grams.
I will likely go back to using large syringes for smaller quantities of putties.
Rolling by hand can work really well with fully sized gloves. I find the little finger gloves aren't great for rolling into rope like strands.
Should you not pre heat putty before applying so it would be even more compressed?
I haven't tried preheating it. It compresses fairly easily, and I haven't had core contact issues (with the exception of that U6 Pro video where I used WAY too luch putty). The putty itself doesn't conpress like a pad does.
It would be interesting to see if it flows easier or not when warmed, but I don't think it's designed to change viscosity when heated.
Just wanted to say that your channel is great. Honestly far better data than a lot of really popular channels.
I'm looking to address some thermal issues on my XPS 7590 - or at least mitigate them a bit as it's getting quite old.
I'm going with PTM7950 for the CPU & GPU but I'm also debating if I should change the stock thermal pads. Is putty tangibly better than after market thermal pads?
Some years back I did a mod where you basically stack thermal pads to dissipate some heat from certain MOSFETs to the chassis. It did help but I wonder if putty would be more effective than stacking thermal pads to make up the height 🤔
Thanks for stopping by and your kind words. I think putty will worl well for you. I have an older chart comparing putty/pads/shims located in the VRAM chart section at the link in my channel about page. It's now obsolete since I wasn't tracking ambient back then, but it will give you a rough idea of relative performance between ourty and pads.
I'll be retesting pads with my new setup when I can.
@@snarksdomain PTM7950 has now arrived along with some CX H1300 13.5W/m.K (seems liked a good compromise with cost vs performance for my budget). Do you have any suggestions or rules of thumb you use for amount to apply? I'll probably just be using some gloves and rolling it by hand.
Edit: Just found your response on your most recent video with instructions. Will follow those, thanks again!
What is happening with the core temps? They are going up as the memory goes down.
@gumbislice1605 that's what I though too, but I wasn't getting good enough compression. In the future I will opt for thinner shims to make it easier to compress. If you take a look at the most up to date charts you will see there isn't a corelation between mempry temperatures and core temperatures. This is because the each VRAM only produces around 3-4 watts.
The charts are located at the Google Drive Link in the video description (Test Chart Repository>VRAM TIM>ETCHASH).
@@snarksdomain that's makes sense. I'll take a look at the drive. Thanks for getting back to me!
Hi Sir can u tell me i use to replace rtx 3080 ti gaming oc 11gb,i will buy 20gram,50gram or 100gram for 11gb vram
Your 3080ti should have 12GB of GDDR6X. If you only want to put putty on VRAM then 20g should be enough. On my test card I use around 14-15 grams on 8 VRAM (3070 ti).
If you want to do putty on whole card and replace all thermal pads I recommend getting at least 50g to do the job.
You will want to check the latest charts located on the Google Drive (Test Chart Repository). There is also a putty calculator there that you can use to enter your OEM pad dimensions and it will help calculate the amount of putty needed, and also the diameter of the logs to roll (if you use log method).
Ok i am wrong 12gb 😁.tks u so much sir
@QuyLePC no worries :) if you run into any issues, a great place to get help quickly is on the Discord server (link in video description)
@@snarksdomain you are so kind.tks alot 👍
@@snarksdomain Hi sir can we use thermalputty for VRM phase?or we use only for Vram?
Did you already test the Ultimate version they have now? Gonna try that. Nice vid!
To my knowledge, UX Pro Ultra is the current top performing putty from Upsiren. That is the putty tested here. Is there another putty you've come across from Upsiren? (Other than U6 Pro and UX Pro)
@@snarksdomain Ah yes im blind, thought this was ux pro but at the start it shows ultra on the package. Ty.
@AzeroxDoomraider To be fair there isn't a "UX Pro" that doesn't have the word "Ultra". There is only a "U6" and a "UX" putty from Upsiren. It does make it a bit confusing that they both have the word "Pro" in their names.
Hi Mr. thermal paste and putty master. I bought hy236 putty 20 gr for laptop do you think it will be enough? (with old stock putty). unfortunately AliExpress doesn't send putty's to my country so i couldn't buy U6 Pro
i don't game by the way
@gibonzproject9710 Nice. 20g is enough for most gaming laptops. I think you'll be happy with HY236.
For Upsiren putties, I usually go with UTP-8 now.
@@gibonzproject9710 Does your laptop have a GPU? What model of laptop do you have?
@@snarksdomain ryzen 7 5800h & RX 6600
@@snarksdomain no issues on gpu VRAM right now but old stock putty dried out a lot (something like K5 pro white)
new sub, relaxing to watch
Will 20 grams of this putty be enough for one video card? (rtx 3080)
@silvatis8061 You'll need around 50g for a 3080. I'd likely go with Upsiren UTP-8 or Fehonda LTP81 these days.
I used 44 grams on a 3070. That had 8 VRAM chips. Your 3080 will either have 10 or 12 VRAM chips.
I was wondering how much I would need for my 4090 FE, is 20g enough?
@shinotabi 50g for most GPU's. 75g for a 3090. 20g for laptop.
Be sure to check the most recent charts at the Google Drive link in video description (Test Chart Repository > VRAM TIM > ETCHASH)
@@snarksdomain thx 🙏🏽 oof that’s gonna hurt. Around 100€ in Germany for UX pro 40g.
@shinotabi that the wrong UX Pro.
Maybe consider going Upsiren UTP-8, Fehonda LTP81, or even Honeywell HT10000 from. EUGENEY_KH (youtuber).
Even CX-H1300 13.5w would be a good option.
You can buy Fehonda pitties directly from the ma factures via their website contact. Same woth Halnziye putties. Also Laird T-Putty 607 is another good option (Moddiy)
Or see if this listing ships to Germany? www.ebuy7.com/index.php?route=product/product&product_id=479964582804248
You could even go with Halnziye HY236 or HY256 directly from Halnziye.
Hi, what are your thoughts on only replacing the vram pads on a 3080? Would it affect the temperatures if i only put the putty on the vram while the vrm and mosfets are the original pads? The reason i do not want to change them is because i read the 3080 cards uses other pads as spacers so having the right height is very important. Also my card is a very obscure card with not a lot of info regarding pads and in those sources people are claiming 1 mm pads and others 1.5 mm, so i just dont want to deal with that if i dont need to. Thank you in advance for all your work and efforts youre doing here, you deserve a lot of recognition for the quality of your content.
Thank you for your kind words.
Which make/model of 3080 do you have.
I don't see an issue in only swapping VRAM pads out. Generally those are the ones that bake/leak oil sooner anyways. And even if you decide to change everything out later, you can just reuse the putty by reforming it by hand, at least for a reasonable amount of time.
I regularly only swap out VRAM interface materials during testing. In my case, I'm using TG-PP10 putty everywhere else (and MX-4 on the die).
We won't know the long-term performance/reliability of UX Pro for quite some time, but I'm fairly confident it's going to be as good or better than most pads out there.
@@snarksdomain Thank you for your reply. I keep replying but my reply is not uploading for some reason. Anyways, I have a gigabyte turbo 3080. Its a blower style, so it gets quite hot and loud at times. I am hoping this putty along with MX-6 will fix that. Ill be updating my results! I hope this putty serves well for many years to come with no oil leakage. I'll be buying a depth gauge and using it before putting this putty to determine once and for all what thickness i would need in case i need to change the other pads.
@Tree Top Interesting that it's a blower style. This Putty should definitely allow the memory to run cooler and hopefully slow the fan down enough for you to hear the game you're playing.
I usually save the stock pads by putting them on some parchment or wax paper, folded like a book, and then into a zip lock bag.
@@snarksdomain It really is an interesting card. It is extremely small for a 3080. Nvidea have blocked production of any geforce turbo card because of "cooling", but considering the rtx30 cards horrible cooling, the real reason is probably for the same reason they cut off SLI, they seem to be a good replacement for their overpriced workstation cards. I noticed in a video you said you live in canada as well, do you take donations through interac etransfer? i dont know how to use crypto
Com licença uma pergunta, com seu conhecimento de pastas térmicas qual pasta térmica você indica para reparo em celular, por que estou com um poco x3 pro problemático tenho o cooler DL05 mais não e suficiente, agradeço a atenção e prosperidade no seu trabalho
Olá Victor. I watched a bunch of videos and read posts about that phone. It seems like many people have thermal issues with it. How to fix those issues depends on how experienced you are with phone repair, and how much modification you want to make.
It looks like there is a gap between the processor and the heat pipe, where Xiaomi used some thermal paste, but not enough. It doesn't get fully compressed so the "thermal bridge" provided by the paste is not good enough. You could try just cleaning it off and applying more paste to make a better "bridge" for the heat to get to the aluminium frame. If going this route I would probably recommend PTM7950, from the syringe. PTM7950 sheet won't work well in this application as the gap is wider than the sheet is.
Another option would be to try just cleaning off the paste Xiaomi used and apply a putty there instead, like U6 Pro. I would go with a sticky putty on something like this I think. UX Pro is more clay like.
If you wanted to go further and take the copper foil tape off, you could apply some thermal paste, or PTM7950, between the CPU and the foil, then replace the putty that is on the memory chip with more than Xiaomi used (maybe try Upsiren U6 Pro). You could even use PTM7950 between the CPU and foil. I would only do this if you have access to copper foil tape or if you can peel it back very carefully without tearing it. Then thermal paste/putty/ptm on top of foil again.
I've never disassembled a phone that far yet myself, but maybe some day :)
@@snarksdomain obrigado sério muito obrigado mesmo eu ando procurando respostas sobre isso a dias já que é um assunto pouco comentado, enfim vou usar seu conselho muito obrigado mesmo pela atenção desejo muito prosperidade no seu trabalho💚
My pleasure. For your reference, here are some of the videos and posts I read/watched about your phone to get an understanding of the issue:
forum.xda-developers.com/t/thermal-pad-copper-mod-for-poco-x3-pro.4479421/
ua-cam.com/video/1Tkmqn2fFMg/v-deo.html
ua-cam.com/video/Ixl8tkhLLjk/v-deo.html
ua-cam.com/video/tXKVTUt5VHE/v-deo.html
imgur.com/a/wCq5IPK
@@snarksdomain nossa obrigado novamente,eu nunca iria achar esse conteúdo obrigado muito obrigado mesmo deus abençoe sempre 🙌
How is the longevity when comparing with PTM 7950 longevity on laptops?
I wouldn't pick UX Pro for a laptop anymore. I'd pick a softer putty.
As for longevity of the different thermal putties, it will be a while before we know for most of them.
TG-PP10 definitely set the bar high. I recently opened up my 3090 which had copper shims and TG-PP10 for 2.5 years. The putty was still soft.
Here's hoping some of the other thermal putties can match that.
@@snarksdomain Do you recommend Upsiren Utp-8 for laptops and how is the longevity of it?
@mehmetoz1932 it's one of the putties I'd pick for a laptop based on it's softness. It's far too new of a product to know it's longevity, but it was designed to have better longevity than Upsiren U6 Pro, according to Upsiren, and there are folks who have had U6 Pro in their laptops for a little over a year now.
Honeywell HT10000 seems to have a good track record of 2 years so far, according to folks buying it from Eugeney_KH (UA-camr). That's where I got mine from.
I also consider Laird T-Putty 607 to have a good track record thus far, as it's seen a bunch if use in Europe/Germany.
Halnziye is another well established manufacturer of putty and really competitive pricing as well. You can actually contact Haknziye directly using their website and purchase from them. Their HY236 and HY256 and an excellent putty for laptops as it is very soft. It'a performance is more than adequate for laptops. It wouldn't hesitate to use it.
Penchem's TH930 has the closest resemblance to T-Globals TG-PP10. Pemchem is also a well established Thermal Putty manufacturer and has their products listed on Dogikey even (depending on which country you're shopping from).
Fehonda has been making Thermal Interface Materials for quite a long time as well. It will be interesting to see how their LTP81 and LTP65 hold up over the coming years.
I don't know what the best choice of a putty is for you. A lot depends on where you live, which can limit your options (depending on country). It also depends on your budget and your willingness to try something new.
If you want top performing, worh a decent track record, from a large manufacturer and good softness then I think Honeywell HT10000 is the way to go. We're already trusting the longevity of PTM7950 from Honeywell, certainly they know how to make a good Thermal Putty.
@@snarksdomain Thanks for the informations. My device is stock and under warranty at the moment. How much temp differ for a 100C CPU in torture test?
@mehmetoz1932 I don't really have a way of answering that question. I have 3 laptops. A Clevo P170EM, an Aorus 15P YD, and a dell laptop for work.
Your best bet is to search forums or discord channels for your specific laptop model and see what others are getting wirh Honeywell PTM7950 on the CPU/GPU core and putty everywhere else. The putty doesn't cool the CPU, but it does allow for more CPU/GPU core contact/pressure to give the potential of a very thin layer of Phase Change.
Hi Snarks, thank you for your testing, found you on reddit and you’ve been very helpful. Might need your advice with something, i want to change the thermal putty on my ps5 , a while ago you were recommending Th 949 1, what do you think about upsiren u6 pro or ux in this case ?
I like all 3 of them to be honest. Whichever is accessible and fits your budget will likely be fine.
I'm really interested in knowing what the hotspot temps are during testing.
For me to track that I'd have to do the testing within Windows. I may go that route and run HWINFO for it.
I can give you the results from my 3090, that I already have installed UX Pro on the VRAM+VRM, and a PTM7950 pad on the GPU die. Both the GPU core and hotspot only dropped around 5-6C (from 72.5C and 93.0C, respectively, to 67.8C and 89.5C). That's to be expected honestly since hotspot is still the hottest reading in the die only. VRAM on the other hand had a huge drop of ~20C. I use Stable Diffusion to stress test the VRAM, and UX Pro dropped it from 104C to 84C. But in any case, using thermal putty is what allowed these hotspot temperatures. I tried pads before, and temperatures always crept up because the pads were never at the right height.
If you want to lower hotspot temperatures, I'd look into undervolting first. On this 3090 I'm running 800mV@1755Mhz and -500Mhz on the VRAM (for my use case, what I really need is VRAM amount, losing about 3% performance doesn't bother me). That, combined with larger fans I mounted, and a some small heatsinks on the card's backplate leaves my 3090 running on full load at 59.1C GPU Core, 75.3C Hotspot, and 78C VRAM. So all in all, UX Pro was very much worth it for my card, and I will very likely always use putty instead of pads from now on.
@@joaoassuncao1 That is awesome! Thanks for sharing your experience with putty. #TEAMPUTTY
hi, great video! can i ask you for some advice I have an asus tuf gaming a15 fa506iu notebook? What do you recommend for VRM, SMD, VRAM? Can I put UPSIREN U6 PRO there? or what is the best choice? Thermal pad 0.5 is probably big? Thank you for the advice
You can definitely out the Thermal Putty of your choice anywhere there was a thermal pad. Lots of good ones to choose from. Have a look at the charts located on the Google Drive (Link in the description).
@@snarksdomain Can I ask you what would you recommend as the best choice for this laptop for VRM, SMD, VRAM? I also wrote to asus support and they didn't help me there! I'm already desperate! I would prefer to put there
Thermal pad DLTECH 15w/mk - 90x50x0.5mm. But I can't find anywhere that 0.5 is not too big. thank you for the advice
@tomastinka4998 I'd either go with Upsiren UTP-8 or CX-H1300.
If you pick CX-H1300 then you need to make sure the pictures in the product listing have a checkmark in the 13.5w/mk box and not the 8w/mk box. They are both called CX-H1300, and the descriptions of both often state 13.5w/mk. The only different is the check mark. As you can see on the chart, they do not perform the same.
@@snarksdomain last question. when you say that you would go to upsiren utp8, why not take upsiren ux pro or UPSIREN U6 PRO? thank you for answer
@tomastinka4998 have you looked at the charts yet? UTP-8 is at the very top, and softer than UX.
do you have any estimate how many grams to buy for a card like 3090?
For a 3090 I would get 70 grams if doing the entire card. You might be able to do it with 50 grams.
Which brand of 3090 do you have? If it's EVGA you can leave the grey putty on the VRMs as it is also a very high performing putty and can easily be reshaped.
I got 50 grams myself. In my 3090 I used the almost entirety of it on the heatsink side of the card, but take note I went quite liberal on the amount of putty I used, I preferred to use a bit more and make sure I had a good coverage, than trying to ration the putty and risk leaving some area uncovered (for the backplate side, the Gelid Ultimate pads I tried initially made good contact, and I saw no point in fixing what was not broken). If you want to do the full card in putty, I'd say grab an extra 20g, or even jump straight to 100g, but again it depends on the amount you need for an application. For instance my 3090 needed 3mm thick pads for the mosfet's backside under the backplate, that's already a fair chunk of putty if I wanted to replace them.
@@joaoassuncao1 thanks for the info
@@joaoassuncao1 Yeah, a 3090 eats up a lot of putty. Might be a good use case for using 2 putties. UX Pro on VRAM and U6 Pro everywhere else. It would bring down the cost substantially.
I'm using the copper shim+tg-pp10 setup on my 3080 tuf and it's been a huge improvement. I have no current intention of redoing my 4070TI but I thoroughly enjoy watching all of your hard work brought to us in video so that for future reference, I'll know what route I want to go and maybe get a stockpile in advance of it. Thank you!!
Copper shim+TG-PP10 is still a great combo and should serve you well for quite some time.
Did you add 2 copper shims between the intermediate VRAM heatsink and the cold plate?
I put one shim in place between the heatsinks but still need to cut a narrow (~8mm wide) shim for the other side to maximize heat transfer "bandwidth".
It seems like Nvidia might have inproved cooling of the VRAM for the 40 series. Ghost was mentioning his 40 series card seems to have good thermals as well. Hopefully you won't need to open your 4070Ti for a long time.
@Snarks Domain I bought a kit that had about 12 different thickness shims and each was almost a perfect fit over a vram chip. From chip to plate, it was 1 shim with the smallest amount of putty holding it in place, whichever shim took up the most space without flexing the heatsink. For the space between the 2 heatsinks, I had to stack appropriately sized shims to equal the space since I had ran out using them over the vram chips. Again, used putty to hold it in place. Usually that was less than a half grain of rice size.
Even with calipers, this was all trial and error and I had to do this 2-3 times to get it all correct. Too thick and flexed the gpu out of contact too much causing overheat and shutdown or too thin and shifted out of place.
Definitely some pucker factor thinking I bricked my card a few times.
Repaste my 7900xt reference card memory junction temperature still reaches 90 degree I think I hit the thermal limit of the cooler itself
@levi smith 90 isn't too bad. Which putty/pads did you use for the VRAM? What's the airflow like of your case?
@@snarksdomain i used the upsiren uxpro putty on the vram and mosfets the mosfets i made a thin log along them and the vram i made balls and squished them down abit to cover the full vram still with enough on to squish down furthermore with cooler pressure and for the die i used honeywell ptm 7950 I am using a sama im01 case its basically a nr200 knockoff with matx support i want to try the morpheus 8069 cooler by rajintek but there is no info for 7000 amd series compatibility
Ever tried "Laird 607 Tputty"?
I still need to buy some.
Ordered. Should receive it in a few days. 75cc should equal 258.75g which comes out at 1.78g/$1 CAD.
The density being near that 3.5 mark might mean it's a good performer as well.
Ive got some tg a7000 on the way 30g tub i will test against this it might be the goat putty
I found out yesterday my order of A7000 was canceled mid-shipment. So I ordered it again but now won't have it for a couple weeks.
Let's us know your thoughts once you test it out. The more people we have testing putties, the better the data, and more reliable the information/recommendation to use putties instead of pads.
@@snarksdomain yeah mine will be with me in the next two weeks I've been told tglobal are UK based but Taiwan distribute so they are good to deal with
Hi. Again an excellent review. Now I want to order one. But how can I tell which one is the original in Ali Express. Can you help me with that? The link in the description section sends directly to the main page. Can you share an original product link? 🙏
When I click on the link it takes me right to the product. I'm not sure what the issue might be. You can also go to the MIddle999 Store and search within the store for "UX Pro". 2 results pop up, both of which are the correct product. One has a lower price, but higher shipping cost so that they work out to being nearly the same price overall.
@@snarksdomain Thanks. I ordered. I hope it reaches my hand quickly. I also have an Msi Ge 75 8se model laptop. When I removed the cooling blog to change the thermal paste, I saw that the paste on the vrrams had not changed at all since the date it was taken. I put the thermal pad of the floor cleaning them, closed the 1mm thick lid and when I worked, the temperature quickly hit 95 degrees. I understood that where there is putty, you need a pad where there is a putty pad. And I think I ordered the best paste on the market.
@@halitay I will probably make a video of changing the putty on my laptop.
I have a Gigabyte Aorus 15p xd-73us224so Laptop. Currently it has TG-PP10 in it as I hadn't known of other putties when I first got it.
@@snarksdomain Are you going to use this paste on your laptop? It's probably the best paste on the market right now, so you're going to use it. 🤔
@Halit I will use this Putty on my laptop for the VRAM, VRM, etc. But will use normal thermal paste or perhaps Honeywell PTM7950 on the GPU Die and CPU Core.
friend, what would you recommend for a gaming laptop ( vram, gpu 2070 q-max) but with alliexpress? I bought a k5 pro but refrained from exchanging ...
UPSIREN U6 PRO??
Hello Adrian. I take it your 2070 MAX-Q has a TGP of 80-90 watts? How are the thermal on it for you so far? I would recommend either using U6 Pro or UX Pro.
I've currently got UX Pro on my gaming laptop (130w TGP 3080), and am using Honeywell PTM7950 sheet on CPU and GPU die. Temps are good so far and the PTM is amazing on the cores. So much better than the dried out Kryonaut Extreme I left on for 1 year. We'll see if the PTM and UX Pro are still gojng strong a year from now.
If you end up getting a large enough improvement on thermals are you considering running a high TDP BIOS for your laptop?
Which brand of Laptop do you have? The folks over on the Legion Discord channel have been modding their Lenovo laptops to good effect using various putties, shims, etc. It's a good place to check out what others are doing.
@@snarksdomain Wow, thanx for fast replay
I have GX701GX woth i7 9750h
I honestly don't know how to check the card's TDP (hmmm i think is above 115w) but the processor is 85W, I just ordered a honeywall and it's already there I wanted to replace everything, including thermopads, that's why I ordered the most recommended k5 pro :/, I will have to wait more than a month for the u6 pro, the u6 pro is currently in stock super promotion so I would choose it ($ 12 for 50 grams u6 pro is probably a great price)
@@123adif123 That is a great price for 50g of U6 Pro.
How much putty do you need? 20g or 50g
I'd pick a different Putty than Upsiren UX Pro these days. There are cheaper and better options. Check out the charts located at the Google Drive link in video description (Test Chart Repository>VRAM TIM>ETCHASH)
For Laptops: buy 20g
For Videocards: Buy 50-100 grams. 50g will do most cards. Some cards like an RTX 3090 will need closer to 75g due tro VRAM on the backside.
Can we use this on SSD nvme ?
Sorry I didn't get to responding until now. I see no issue with using this on NVME drives. Definitely worth a shot.
u6 pro is also good for sure
Yeah, I had originally tested U6 Pro before UX Pro, but that video got taken down.
I've since been retesting all of the putties. Feel free to review any of the live streams to see the application of each one, or follow the Google Drive link below to review the latest charts (Test Chart Repository > VRAM TIM > ETCHASH).
Currently, the 2 best putties are Upsiren UTP-8 and Fehonda LTP81
@@snarksdomain oh, thanks for that information.
Is it good enough to be in Die of GPU?
It isn't the ideal product for GPU Die. For that I would recommend PTM7950 or another Phase Change Material, or normal paste.
I'm not saying it won't work, but using a putty on a CPU or GPU core is highly experimental. I have used TG-PP10 putty on the core of a 3080, and it is still on it. I will perhaps test putty on core in the future, but there are many other videos I will need to make first.
I'll be testing Upsiren PCM-1 at some point as well as Thermal Grizzly Kryosheet. testing Phase Change Materials will take time (likely 1 month per phase change material being tested).
@@snarksdomain Thanks for your hard work and dedication man. You rock!
Considering the price difference between U6 Pro and UX Pro, the UX Pro is not worth it for ordinary user.
A question: How much putty is enough for 1 GPU?
If we compare UX Pro to the cost of pads it's actually fairly comparable and often cheaper since you don't have to buy 4 different thicknesses of pads, but I would agree that U6 Pro is a great bang for buck solution.
Depending on the GPU 20g may be enough, but 50g would be a safer amount if replacing all pads on the card. If you just replace the VRAM pads, then 20g should be okay, unless it's a 3090 with its 24 VRAM modules.
@@snarksdomain just doing my 3060ti both front and back & everywhere (VRMs, coils, VRAM) consumed about 50g, with backplate consuming the most due to the gap there. As for how well it works, well... this GPU lacks any sensors, so i ended up putting some putty on my motherboard VRMs, which do have sensors.
2:00 thank you
Are these putties reusable?
Yes, most putties can be reshaped and reused for a long time.
@@snarksdomain neat, thanks.
Hi this putty works well on acer nitro 5?
@SebasTian-io1gw These days I'd pick other putties. Upsiren UTP-8 has similar/better performance, is softer and cheaper.
If a laptop has gaps smaller than 0.5mm then something like Halnziye HY236 is a solid choice as it can compress down easier than UX Pro or UTP-8.
The best performing putty I know of is Honeywell HT10000. It costs a bit more and not quite as accessible.
Fehonda LTP81 recently came down in price a bit closer to UTP-8 pricing so that is also a good option.
Updated charts are located at the Google Drive link below (Test Chart Repository>VRAM TIM>ETCHASH).
Thank you . I buy 50g of this putty , so you recomend the others puttyes because are much softer than this?. I dont have other alternative to use this on my laptop now :/. You recomend wait and buy the others puttys?
@SebasTian-io1gw it can work well if you've already bought it. I've only tested the Upsiren one so far, so hopefully you got that one, and not the CSGR version of UX Pro.
I recently opened my laptop up after 16.5 months of having Upsiren UX Pro and it was still soft enough to mold by hand. Just slightly dryer than brand new, so the stuff definitely lasts fairly well so far.
Up to you whether you stick with the one you already bought or if you can still return for refund.
Upsiren UTP-8 is like half the price of UX Pro. Laptops only need 20g or less, so you can do a laptop for arpund $11-$15 CAD.
@@snarksdomain csgr is a fake version? . Can you send me a link from alliexpress of utp 8 please
@@snarksdomain I bought the ux pro on aliexpress , middle999 store . Is everyting alright with my paste?
I cannot buy Upsiren UX Pro from Ali, they don't send this stuff to Germany, I don't know why.
Bummer.
You likely won't be able to source Upsiren products within Europe for a little while. There's a trademark war going on..... ua-cam.com/video/xND1726G624/v-deo.html
just bought this
that looks like A LOT of putty, the thermal pads seem much thinner.
@DragonXDrei it was definitely WAY too much putty. I've gotten better at using less. There's a Putty Calculator on the Google Drive link which can be used for calculating putty log diameter based on the stock pad dimensions.
test Laird Tputty 607 ,THX。
I'll order some when I can afford to, along with A7000 and TH949-3. Between the 3 of them I'm looking at $600 CAD.
@@snarksdomain Very expensive. There will also be a Laird Tputty 900
@Zorro歷史名場面佐羅 yeah, for those putties I will have to wait a bit unfortunately.
@@snarksdomain I can provide you with tputty607, the question is how to mail it?
@Zorro歷史名場面佐羅 I've got a PO Box address, if that could work.
Could you do a GPU test with K5 Pro? While it took a long time to reach the 100°C in the older test it would be interesting if that's the only variable that matters. The few times I've seen K5 Pro in real world uses it performed good, but it was never compared to other thermal putties.
Would be very interesting to see.
As mush as I don't want to use K5 Pro on my test card, I probably should test it so we can have better data and have a real world comparison. I'll mostly do it to show why I think it's a bad idea, and save others the headache.
While the 100 degree hotplate test isn't the final word, it has been fairly reliable in showing relative performance of different pads/putties and GPU tests usually line up in the same order of performance.
Don't have any experience with K5 Pro itself (well, other than my DAC/AMP, but that's from Fiio, not Computer-Systems), but from what I've seen around in other videos and forums in the web, the problem with K5 Pro is that it seems to have a relatively low melting point, with some people even reporting it gaining a bubbly consistency when it started to reach temperatures around 100C. There are some reports of good results, yes, but I wouldn't be surprised is those came from low power applications. I'd wager that K5 Pro's thermal conductivity drops like a brick once it reaches higher temperatures.
I can't beat to repaste my GPU
I'm trying to use it for rtx4090.
Can I use it for 4090?
@AAisAA I'd pick a different putty than UX Pro. There are better/similar performing options for cheaper.
According to online conce sus, for a 4090 you want to aim for ~60 Celcius Memory Junction Temeprature for max OC headroom.
So keep that in mind when selecting a Thermal Putty. You'd want to see where your stock temps are and go from there.
Aside from all that, yes, Thermal Putty is great for anywhere there was a thermal pad or other thermal putty. I recommend PTM7950, or similar Phase Change Material, for the core.
@@snarksdomain If you recommend something else, what would be better?
@@AAisAA pricing can be all over the place, depending on where you love a d cost of shipping/import.
Check the chart located on the Google Drive link below (Test Chart Repository>VRAM TIM>ETCHASH)
I'd expect a guy dealing with such tedious testing on thermal interface materials to know the exact model of his computer case with certainty, because surely such a person would care about how cool his PC will run and that would require him to look for a case with adequate air flow, which means he'd have to make some research at least and then by the end of it he'd have the model etched onto his memory
Oh I do. Lol. My gaming rig is in a Lian Li O11D XL Black. Used to run my main PC in a Thermaltake Xaser III Super Tower (Blue). For a moment I couldn't remember if the Fractal Design was a Meshify 2 or a Meshify C.
@@snarksdomain I see. I have a suggestion for a video, why dont you pick any high performance thermal pad brand, and then use pads thicker than what your graphics card requires, and then compare the results to when you use the proper thickness thermal pads. The title of the video would be "Does thermal thickness really have to exact?" or something like that.
I'm thinking that since thermal pads are generally very soft and squishable, the eventual thickness of a thermal pad once the graphics card is assembled would not depend on the original thickness of the thermal pads, as the pressure from all the screwing would most likely squish all the thermal pads to the 'right size' if any of them are thicker than they are 'supposed to be'.
My graphics card is still under warranty and I'm (for now...) a pathetic coward who doesnt want to open up his GPU and 'risk' breaking something, so I cant make that test myself now. Also there is factor of it having taken me a traumatic amount of effort to fit the damn GPU in my case, so I ideally dont want to go through that again. (Case is Meshlicious, GPU is an MSI 3080 Gaming Z Trio)
@Andy96 surprisingly, that's basically how I ended up here. First with trying Gelid Ultimate pads and utterly failing and getting horrible temperatures and core contact issues.
Then I bought a bunch of Gelid Extreme pads because they were supposed to be softer and more forgiving. I had mixed results even with the "correct" pad thicknesses everywhere.
I think the main problem stems from 4-6 pad thicknesses on a card all needing to make contact and still get core contact for it to work out.
The beat chance at making pads work out is likely to compress then by rolling them down to about 105% of the gap size. That way they only need that little bit extra squish from the mounting pressure for everything to make contact.
If you like you can check out some of my earlier attempts at repadding cards with mixed results here www.reddit.com/r/nvidia/comments/pnwogn/asus_strix_3090_thermal_putty_tgpp10_instead_of/?
@@snarksdomain Then I take it that they arent 'that' malleable, because if they were you wouldnt have any problems when you used thicker thermal pads, they would have just comformed to the proper thickness under the mounting pressure. So somehow they have enough rigidity to resist the mounting pressure and bend the cooling plate, which in turn causes different parts of the PCB to lose proper contact with the cooling plate, causing high temps? This is theoretically very stupid and is very hard to believe. How tough could they possibly be? What could I be missing?
EDIT: I'm also wondering, is the reported memory temp pulled from just one specific memory chip, or is it an average of the temps from all the memory chips? This could be tested by cooling every chip manually without the cooling plate...
@Andy96 that was my surprise as well when first trying to repad videocards. I thought that if I took accurate measurements or even followed other people guides wkth listed thicknesses then all would be fine. I bought a few hundred dollar of thermal pads (Gelid Ultimates and then Gelid Extremes).
Extra thick pads won't bend the coldplaye or cooler. They are much too rigid for that. What will happen is you might not have any core contact at all.
Since moving to putty I haven't had to worry too much.....except for when I used WAY too much putty during the Upsuren U6 Pro (About 3x as much as was needed). But there is a much larger margin of error when using putty.
I will add that using pads on VRAM and outta everywhere else can work good too if you REALLY want to use a pad for VRAM. Just gotta be sure to get the correct thickness. But getting one right is easier than getting 4-6 right....especially when some EVGA cards use 2.25mm or 2.75mm pads.
Memory temp shows the highest memory temperature.