More FPS for just $10 - K5 Pro Viscous Thermal Paste
Vložit
- čas přidán 23. 07. 2024
- Try FreshBooks free, for 30 days, no credit card required at www.freshbooks.com/linus
Use code LINUS and get 25% off GlassWire at lmg.gg/glasswire
By replacing the thermal pads on your laptop or GPU with K5 Pro you can get more FPS, but is it worth the mess?
Buy K5 PRO Viscous Thermal Paste
On Amazon: geni.us/Y3e2
On Newegg: geni.us/a3DZb
Buy Alienware M17 Laptop
On Amazon: geni.us/YXgeMa
On Best Buy: geni.us/qtrtS
On Newegg: geni.us/IJqM1
Buy Asus RTX 3090
On Amazon: geni.us/rStWsA
On Newegg: geni.us/dVNSZb
On B&H: geni.us/jfwRRx
Purchases made through some store links may provide some compensation to Linus Media Group.
Discuss on the forum: linustechtips.com/topic/13992...
►GET MERCH: lttstore.com
►SUPPORT US ON FLOATPLANE: www.floatplane.com/
►LTX EXPO: www.ltxexpo.com/
AFFILIATES & REFERRALS
---------------------------------------------------
►Affiliates, Sponsors & Referrals: lmg.gg/sponsors
►Our WAN Show & Podcast Gear: lmg.gg/podcastgear
►Private Internet Access VPN: lmg.gg/pialinus2
►Our Official Charging Partner Anker: lmg.gg/AnkerLTT
►Secretlabs Gaming Chairs: lmg.gg/SecretlabLTT
►MK Keyboards: lmg.gg/LyLtl
►Nerd or Die Stream Overlays: lmg.gg/avLlO
►Green Man Gaming lmg.gg/GMGLTT
►Amazon Prime: lmg.gg/8KV1v
►Audible Free Trial: lmg.gg/8242J
►Our Gear on Amazon: geni.us/OhmF
FOLLOW US ELSEWHERE
---------------------------------------------------
Twitter: / linustech
Facebook: / linustech
Instagram: / linustech
Twitch: / linustech
FOLLOW OUR OTHER CHANNELS
---------------------------------------------------
Mac Address: lmg.gg/macaddress
Techquickie: lmg.gg/techquickieyt
TechLinked: lmg.gg/techlinkedyt
ShortCircuit: lmg.gg/shortcircuityt
LMG Clips: lmg.gg/lmgclipsyt
Channel Super Fun: lmg.gg/channelsuperfunyt
They're Just Movies: lmg.gg/TheyreJustMoviesYT
MUSIC CREDIT
---------------------------------------------------
Title: Laszlo - Supernova
Video Link: • [Electro] - Laszlo - S...
iTunes Download Link: itunes.apple.com/us/album/sup...
Artist Link: / laszlomusic
Outro Screen Music Credit: Approaching Nirvana - Sugar High / approachingnirvana
Intro animation by MBarek Abdelwassaa / mbarek_abdel
Monitor And Keyboard by vadimmihalkevich / CC BY 4.0 geni.us/PgGWp
Mechanical RGB Keyboard by BigBrotherECE / CC BY 4.0 geni.us/mj6pHk4
Mouse Gamer free Model By Oscar Creativo / CC BY 4.0 geni.us/Ps3XfE
CHAPTERS
---------------------------------------------------
0:00 - What is K5 Pro?
2:19 - Stock Test Results
2:38 - CPU & GPU Thermal Paste replacement
5:37 - K5 Pro application
6:20 - Why K5 Pro thermal paste is better
6:48 - Interrupting Linus goes?
6:55 - How do you clean K5 Pro off?
7:17 - K5 Pro application continued
8:33 - Repaste & K5 Pro Thermal Paste results
10:45 - Outro - Věda a technologie
You guys need to do a "ultimate thermal performance" build with all the ingenius things weve seen in the past. K5 Pro, the black nano cooling fluid, Kryonaut, just everything.
Go all the way. Submersion cooling, but replace the mineral oil with ALL LIQUID METAL!!
(Yes, I know that would eat your computer)
Don't forget the custom cooler
@@alexmcd378 Actually, it wouldn't get a chance to. As soon as you turned the thing on, that super-conductive metal would fry every component everywhere.
TO THE LABS! _Batman tune_
Yes
I would love to see a review of this with more midrange components, maybe a GPU that doesn't have any thermal pads to connect to the backplate, then comparing this to some cheap aftermarket pads.
This! With overclocks!
Highly doubt if you'll see bunch of reviews... Heck this thing existed for almost a decade and you only found it because of this promotion
Wouldn't really make any difference in a card that isn't thermal throttling.
I second
Maybe you could apply this in a laptop to connect the heat sink to the chassi, like they did with the macbook a while back
The gain on the 3090 repaste was actually over 20% in the 1% and 5% lows in furmark which matters way more than the average. Btw thats an enormous gain
which is good, until heat transfer through the goo to nearby less heat tolerant components will eventually cause them to fail.
@@HG-fw8hn that's a really good point I have never even considered before.
@@HG-fw8hn Not really
@@PaffDaddyTR Not really what? The purpose of this stuff is literally to transfer heat. For instance in a Laptop the thermal solution is carefully calibrated to transfer the amount of heat from components that comes from a thermal pad cooling solution. If you slap this stuff on then:
1. Heatsink becomes way over saturated from heat, which it can not dissipate fast enough. Which will probably lead to more thermal throttling of CPU/GPU/Memory
2. All that heat that the Heatsink can't get rid of gets saturated to nearby components that are covered in 'goo', through the 'goo', which will increase thermal stress on them well over operating limits.
It's simply thermal dynamics
@@HG-fw8hn Thermal sensitive components wont be laid right beside the VRMs, CPU and memory, since heat transfers through the copper lined PCB much more efficiently than through thermal goo, bud. And cooler "oversaturating" with heat? Seriously? Can you name an example of that happening?
Quick story in regards to removing batteries before disassembly: I forgot to remove it a few weeks ago on my ASUS laptop (ironically, en route to replacing my thermal paste) and, in the process, accidentally let the shield covering the display driver touch a ferrite bead on the board. Unfortunately ASUS decided that, despite the machine being off, that connection should still have a fairly uncontested live connection to the battery. In less than a second, an impressive arc formed and quite literally blew the ferrite bead straight off the board. Cue waiting a week or so few a strip of ferrite beads, a hot air reflow station and some chip quik all because I forgot to unplug the bloody battery
Hmm the worst part is the battery could have been damaged
@@c4_yrslf726 If the connection is held long enough, I definitely agree. I don't think a very quick short is going to do enough damage to worry about, but if it made prolonged contact that would absolutely be a massive fire risk - particularly with a LiPo battery. My biggest concern would be catastrophic component damage - nothing like feeding uninterrupted current and an unknown voltage through your shiny RTX 3080
To clarify, are you talking about an RF shield? Those should normally be grounded
@@__lasevix_ It might've been an RF shield but I think it was to act as a barrier to stop any loose components shorting across the LVDS connector. The problem was actually that it was grounded lol - the ferrite bead had an open positive connection, but a closed negative one so when it tapped the shield it formed an open circuit 🤦♂️😂
@@Jacobwlane oh okay. well that's a good example of why you should unplug the batteries first
I've applied K5 pro multiple times, and even with alcohol is a really BIG pain in the *** to remove it. Maybe I'm not the best at removing it, but either way, it's not as easy as they say. And because you need to apply a lot, so it can fill the air gaps, when you put back the heat sink, the paste overflow to the sides, making a big mess. It's not conductive, so its not really a technical problem, but it looks like a big ugly mess.
What do you mean? I like the look of a vanila ice cream GPU sandwich. :)
I recently used SYY 15.7W/m.k paste and it's like concrete but was the difference between throttling and not throttling with an h115i and a 5950x.
And then I switched to a 150i anyways but hey.
I'm debating sticking some on my 1080ti and was wondering about the pads, so this video and all the comments is great.
Are all pastes non-conductive? I got some kryonaut on my board and freaked out about cleaning it.
@@americansmark They *should* be non-conductive, but better safe than sorry.
You can say ass on CZcams. Ass.
This isn't really surprising: it's essentially a form of thermal paste, and thermal paste is better than a thermal pad. It would have been more interesting if you compared smearing an existing thick thermal paste, to this K5. Also there is the question of how it ages, does it dry and crack?
Thermal interface material has to be thin to do its job best. Normal thermal paste is often too runny. Except you use Kryonaut Extreme. But that stuff costs 99 USD for 33 grams. I think this stuff is designed to have a high thermal conductivity up to 2.5mm which is the thickest pad thickness widely used.
Just do what he told to do use termal grizly than cupper than thermal grizly and you have done it...K5 seems to be advertisment stunt than a real deal...
If left sitting in the tub, the outer area does. But that was only for the one I had previously opened. The other unopened one was fine. For me that was about 3 years. Don't know about in the system/daily driving though as I never fixed the iMac. I will eventually though as I'm now using it on an old server motherboard which had a very worn thermal pad on the CPU VRMs.
Found out about this stuff when re-pasting an iMac GPU and it turns out Apple uses this stuff from the factory. Couldn't fix it though.
@@alvydasjokubauskas2587 Did you actually watch the end of video? That's not what he said.
@@alvydasjokubauskas2587 That requires precise machining of the copper shims, though. If you use shims too thick you'll fuck with the mounting pressure on the CPU/GPU.
This is really interesting stuff. I've been resigned to keeping a bunch of thermal pads of different thickness around and taking care not to use the wrong thickness. This will definitely take some of the headache out of custom watercooling future GPUs when thermal pads aren't pre-cut and pre-applied.
I bought this stuff right after watching the video and used it and mettalic thermal grizzly on my strix 1070 thats been running almost non stop since i built the pc in 2018 and i went from 70fps to 120 in gta 5 and from 52-70 in cyperpunk
@@andreaskeltwolfe6176 respect
Been using K5 Pro on GPU VRM's on Laptops especially for 2 years, no issues so far and good performance. Much easier then trying to figure out the pads thickness on each laptop.
Chupapi munyanyo
@@abc20243 For what reason? There's no point in using K5 Pro on an SSD, Use thermalpad instead.
@@abc20243 you can use it with no issues.
This was great, would love to see it on someones older daily driver gpu.
@@louis__________________8281 shut up bot
@@cherrikola999 Report it.
@@Ryuseigan done
I just bought some. Going to use it on my 1660 Super. Over clocked.
@@Ryuseigan YT doesn't care. I am reporting all of them, none ever get removed.
It would probably be pretty easy to apply if it was in a piping bag.
How about in a syringe.
@@matthewtremain683 Only Fauci can apply it going forward.
I read online about and then used K5 Pro a few years ago to replace the underperforming thermal pads of VRMs on an Asrock 970 Fatal1ty Performance mainboard with an FX 8320 on it. Used to be my main desktop until early this year and is still stable with a mild overclock. I was also pretty proud to discover back then that this product hails from Greece. :)
Another cool use-case for this would be heatsinks for M.2 SSDs that use different height ICs, such as the V-NAND ones, where the controller is a good 0.5mm shorter than the memory ICs
My question is... what's its usable lifespan? I might want to clean it off to apply fresh goo. Or are they implying that it doesn't become less effective as time passes with heating and cooling cycles like traditional TIM and pads do?
I'd guess it lasts about as long?
Bit of a difficult thing to test for them.
It's not really a goo, it's a molding pad that turns into soft solid. It can last for years.
@@jort93z That was my assumption as well.
@@mikeycrackson Ah... guess I should stop being lazy and just gone to their page. Interesting though... I have my doubts as well.
@@Ryuuken24 I was using "goo" as a colloquialism. It was used for slight comedic effect. I was not being literal.
I replaced my thermal pads on my old 1070 Ti Duke a few months ago with cheapos I found on Amazon. Both my stable GPU and Memory overclocks dropped dramatically. 250 for memory/90 for gpu. So I ordered some of the 6 w/mk Fujipoly pads, and redid. Memory offset went up to 400, 50 higher then stock, and my stable GPU offset went up to 130, 20 higher then stock. The same thermal paste was used both times, and the average GPU temp was roughly the same around 54 degrees. Just shows that good thermal pads do make a difference, and you can't always tell by just the GPU temp itself. Although in the grand scheme of things, the difference in offsets vs stock might equal 5 fps, if that.
Get some good stuff, like the 12.8 W/mK Thermalright Odyssey. That will make a huge difference.
@@matasa7463 I could've gotten 13 w/mk pads, but the difference probably wouldn't be that dramatic. Maybe an extra step on my OC.
Trying it the cheap and lazy wsy, I bought a sheet of Siilpad, nasty cheap stuff had no glass in it so was insulator, 100C CPU temp and 60 idle, replaced with MX4 and got 20C 40C. FX-6200.
@@joefish6091 who uses pads on a cpu again?
@@KryGaming2k Also, always so interesting when people get sub ambient numbers on their CPU.
Cool how so many is able to break the laws of thermodynamics!
I've used this and it works really well. I live in a hot city and the K5 pro got rid of all the persistent thermal throttling I was experiencing.
I remember that the old Artic Silver directions had you smear some paste on the copper block of the cooler, then sort of polish it into the surface of the block. I can remember testing my 939 Opty and saw a good benefit in the practice with my Danger Den TDX. Filling in those gaps is important!
We all know that RGB increases performance by at least 120%
Lmao true😅
That was funny 3 years ago
No, RGB improves performance by 1000%
No it does not
Only red does, blue drops the temp by 20 c.
This was great timing...been planning on pulling my laptop apart for a deep clean and repaste, so I'll be grabbing a bit of this beforehand.
I used thermal putty on my gpu, stock zotac 3080 amp was improved a lot with a double sided water block, adding the thermal putty later on made very little difference to memory temps and performance but it was a little better.
My expected gains are in the vrm’s and longevity.
I'm using the K5 pro for a few years now. Used it also successfully to paste a throttling NVME SSD on the underside of a motherboard. Pasted it to the case actually, worked extremely well.
So you basically changed the VRM and memory thermal interface but didn't offer before and after temperature results?
I don't understand why, the FPS numbers are in the margin of error. Measuring the temperature at the same fan speed and ambient would show an objective difference, if any. Kind of disappointed on this one.
But keep up the good work, I expect better next time. At least from LTT.
yea, doesnt use temp is really weird. Maybe it's because they use it for vram and older vram before gddr6x (which mostly of them) doesnt have thermal sensor so there is no what to tell... But still, they use it for CPU too...
I really liked when it was mentioned that "GPU still needs to run 3D Mark *once*". ONCE? THAT is the benchmark they are doing? If just one run can prove anything, I might even have had an overclocking record somewhere in the past... The lack of depth on the really important things on the videos is abysmal... :(
Came through to check out LTT, it's been a while...
Another sad and typical video with no value. No numbers at all. Yes technically they were answering the "claimed performance gains" but not the actual figure we wanted.
There is always so much potential and they never fulfill it which is why I stopped watching LTT for the most part several years ago.
Another one of my pet peeves was the Morpheus II cooler review. They literally said nothing about it that would have any value. That cooler changed my GPU entirely as I got on sale for cheap a GPU but it had a blower style cooler on it. It was whiney and ran hot so I looked into a solution. Still cheaper in the long run but the Morhpeus II made it dead quiet and ran 25-35 degrees cooler at stock... Allowing for some overclocking as well at the same noise floor. But they didn't cover any of that in their review, nor any numbers just "the fans look like they are running at low RPM and it's quiet"...
It might be easier to use if they sold it in a syringe like thermal paste with a flat nozzle to squeeze it out in layers.
A caulking or grease gun might work better since they are intended for substances with a high viscosity. And if you get an electric grease gut you bight be able to mod it so the grease comes out slower for easier application.
For older systems that already need new thermal paste, this is a great idea. Sure the gains aren't *massive*, but if you're already tearing down the system, might as well tear it down and get rid of the old thermal pads/paste too for some extra gains and (potentially) increased longevity of components.
Hello fellow Tim!
I'm glad to see the results for this interesting project but it must have been pure hell to go through. Good job.
10$ that’s 10 McChickens 😬
bruh. true
Cringe “subscribe to me” banner fake account
McChickens are $1.15 where I live :(
In the last few months the price of a McChicken in my local McDonalds shot up from 1$ to like 2.20$.
@@clawdroppin same
How long does this stuff last when applied on a component? I saw one review on Amazon claim that he got this stuff replaced by the company because his arrived drier than usual. If it dries even before it's applied, what is its shelf life too?
Could just be a bad batch, rather than it having a bad shelf life or drying before it was applied. Hard to say, since they do claim it has infinite shelf life and can withstand up to 240 degrees Celcius. Grain of salt ofc but unless that is super common, could just be a rare occurance.
They are still at startup phase it seems, so quality control will be poor for a little while before they mature or outsource the production to a reliable manufacturer.
Welp, I tried using this stuff instead of thermal pads on my 3080 since you guys suggested it and my vram junction temp went up 12c. I opened it up a second time and added a bunch more since I thought maybe i hadn't used enough the first time. After the second attempt, I was still up 10c from where I was with the stock pads.
Good call LTT! I'll bear this in mind for my next build. Thanks for the info!
I've been using it since last year for every laptop repairs and some smartphones who use their own paste, and it works fine. You can avoid having a bunch of different quality and thickess thermal pads, and the repairing is more easy and reliable
A lot of claims out there saying the K5 can boil .. have you seen this?
@@timotmon never happened to me
@@carlosgomez_c137 Thanks Carlos!
I've been considering repasting my RX580 with something like cryonaut, and knowing I can replace whatever thermal pads it has while I'm at it is pretty nice. I will almost definitely be picking up some of this stuff.
I repasted my rx570 with kryonaut, dropped about 10c, at full load used to be 80c-85c(depending how hot the room).
Kryonaut isn't designed for air cooled GPU temps long term. You will have very good temps initially, but expect it to creep up over time. Gelid Extreme is very stable for GPUs, that's my recommendation. Having used K5 Pro, I also wouldn't advise it because it leaches silicone oil like no other. I rtepalced my mobo VRM pads with it, and after about a month, had streaks of oil running down the entire length of the mobo and under the PCIe slots. It was a huge mess and a pain to remove
@@taylorsharp5928 So TDRL: Use Gelid Extreme for everything, including as a replacement for thermal pads.
@@taylorsharp5928 it's been a year now, no issues.
@@GENKI_INU can confirm. i have gelid extreme pads on my reference rx480 and liquid metal on the core. they work fairly well and the best i've tried on this very hot card
I'm playing around with this and scrap copper for getting my gaps for cooling down to like 1mm and using the goop to fill in the blanks.
Something I've noticed is that many of the components that get cooling pads have some extra wiggle room in how they are mounted to the board. On casual distant examination they look very flush, but when a flat surface is applied to the tops of them there is some very obvious tilts going on on some of the components. Which is only amplified when circuit board flex happens from mounting hardware to it.
So right now my target is using some thin scrap copper to get everything down to a 1mm or less clearance, using a little fine sanding (with a vacuum to pick up any particles produced, and remember for this less is more !!!!!! )) and then using this goop to fill in all the blanks when the heatsink is reapplied.
Best New Year's video on Tube! Thanks Alex and LTT for opening my eyes to this new compound! I have a dented heatpipe on a Gigabyte VGA card (that resets when is 90% loaded because of poor contact with the chipset). I will try this, for that dented spot! With MX2 filling the dented spot is resetting the system.
Honestly, I really missed a thermal test. I would love to know, how much cooler the VRM was.
Like they said, there's no temperature probes for the VRMs. They *can't* measure it, unless they had the means to probe the inside of the chip (Which is very difficult to do manually).
@@dizzydaisy909 then they should've used a card that does work on... Works on my 3080 FE!
Where's tech Jesus?
@@dizzydaisy909 Exactly like Bliznade said. Or they could use a thermal camera for the GPU or use their own thermal sensor as Hardware Unboxed uses for their reviews.
@@EaZynite I agree. FPS numbers from a CPU bound game was not the correct tool to measure the performance of this stuff. Pretty disappointing from LTT. I mean sure, it's free but why waste everyone's time when they could have actually given us good information with less time necessary on their part? A 3 minute mining test with hwmonitor could have told the whole story.
I bought that for my watercooled gpu. Standard pads were 0.5mm, but the memory chips were not level to each other and I had shutdowns. Applied this stuff and havent had issues after that. Some users reported drying up with time, but we will see. I could recommend for similar issues, but not if thermal pads do the job.
Interesting. Yeah I think it might dry up but it shouldn't be a problem since it should stay fine in the center where it matters. How are your temperatures? For me this video came a week too late. Put on my EKWB Vector Strix block and the backplate block onto my 3090 with the stock thermal pads from EKWB. Runs really cool. Memory ever peaked at like 54°C. And 24GB of GDDR6X is a fusion reactor...
@@kyoudaiken well, my gpu is ancient r9 290x and doesnt have memory IC temperature. Considering that 4 of the memory chips didnt made contact with the block at all and i had to play for 1-2 hours before a shutdown, the K5pro is doing its job now.
@@sreflectionbg excellent I've been looking for a comment on how it works with older cards, I too have an R9 290 so you think it's well worth giving this stuff a go? When I last stripped down my card I noticed the thermal pads were looking worse for the wear and some parts tore when taking it apart. Hoping this goop might extend its life a little more.
I just put my 3080ti on a water block on Christmas, my memory temps went from 98-103c to 55-60c under load. Doubt this has any extra benefit when water cooled.
@@ascap2854 Same here with my RTX 3090! GPU-z shows over 200 watts memory power consumption when running blender. Try to dissipate that with thermal pads... They REALLY need to rearrange the cooling. IMHO, air cooling as it is currently designed does not work anymore when a card takes more than 200 Watts overall. With air cooling you hit a brick wall at around 250 Watts. From there, you need to brute-force airflow. GPUs over 200 Watts are incredibly noisy.
Thanks for amazon link, this is much easier than the thickness+cutting thermal pads hassle
I've actually used this before, not for replacing thermal pads on PC parts, but for re-pasting a Nintendo Switch after taking it apart to replace a broken screen. Nintendo uses a similar viscous thermal paste between the CPU/GPU chips and the EM shield (which also pulls double duty as a heat dissipation plate), and K5 Pro is pretty much a direct replacement for Nintendo's stock paste.
Seeing this done for a brand new ek waterblock install would be interesting. Try timing installation vs the cutting and placing of pads, and quality of final results.
I just ordered some. I most likely will slap this on my heatkiller'd 2080ti, alienware laptop, and my brother's 3080 ti strix. Time to do some baseline benchmarks first
I got legitimate chills when you popped that board out of the gpu.
well one of the most annoying parts of replacing thermal pads is getting the thickness wrong (or just not having that one particular thickness). I dont change thermal pads enough to have a collection on hand so this makes a lot of sense in terms of convenience
I used this on my Asus Zephyrus G14. Works great and it's very easy to install. Along with using Gelid GC Extreme on the CPU and GPU it dropped temperatures significantly. I've got under 80c across the board.
You missed the most interesting point on the 3090 : the VRAM temp.
It is super toasty with stock thermal pads, that would have been great to know how much VRAM temps changed with the paste
I have a video on my channel where I replace the thermal pads and I can improve the temperature from 96º/100º to 68º/70º on my RTX3080 (the video is in Portuguese but in the description I have the video in English for anyone who wants to)
I do not understand what is the point of this video without temperature measurement🤷
@@1vend7 Why is your GPU so darn hot? Ive never had it reach those temps. It should thermal lock or crash when it hits over 70' C
@@alexlabmonkey the point is that you can buy that or similar product and forget abut taking measures for thermal pads and avoid accidents such as putting 1mm thermal pad instead of 1.5mm pad and possibly damaging your GPU.
@@exterminater267 He's not talking about GPU temps, he's talking about VRAMs, those are normal temps for VRAMS
Might be a good thing to test in ltt labs for a few years. Also I'd like to see you clean up the laptop and put the good stuff back in if possible
I think the most important thing about it is... about how easy to is to apply, instead of searching and cutting the right size thermalpads.
However, I do agree with some guy from the comments, that It would be nice to compare how good they are vs. expensive termopads, plus to test this thing on the middle-low end graphics cards/laptops.
I have an Acer Predator Helios 300 (2019) and I just redid my thermals. Used Noctua NT-H2 on the GPU and CPU, and the thick K5-Pro there on the VRAM and other areas were there was that dry pink crap or pads. My temps decreased by about 8-10°. Well worth it.
how could you not include memory temperatures for the gpu? That's like the only reason I would want this stuff with how hot my 3080 memory gets, and you guys didn't even mention it.
I want to see the actual VRAM temp differences on the same fan curve. My 3090 was hitting 115° on the stock pads, and swapping them for Odyssey Extremes brought it down to 72°. If I can see evidence of this putty outperforming other thermal putties and my current thermal pads, I will try using it when I install my waterblock instead of getting an active backplate. It's thermal conductivity is 1/3 of the thermal pads I used, so I'm fairly skeptical
Have you considered copper shims and thermal paste?
@@ffwast that is so underated. I combined it with thermal pads on my backplate for a short test and the heat transfer burned my hand, but saw no temp drops. but I am sure on the memory where I placed it there is definitely a lot of transfer, better than whatever thermal pads are on the market right now.
@@ffwast I have a video on my channel where I replace the thermal pads and I can improve the temperature from 96º/100º to 68º/70º on my RTX3080 (the video is in Portuguese but in the description I have the video in English for anyone who wants to)
Although the k5 pro's thermal conductivity is under half of the odyssey extremes rating (like 5 vs 12), if you take into account the contact ratio of of pads vs paste (60 vs 99) and the garbage thermal conductivity of air (.025) is not a stretch so see how this particular paste can beat high end thermal pads. Though of course proper testing is warranted lol.
@@hjorthrimul4350 that's why I wanted to see the actual temps! It'd be $50USD to get enough to redo my card, which isn't terrible...but like Alex said, how much of a pain would it be to remove if I don't see an improvement?
I bought this last year for my laptop, it did a really nice job. It had something similar already so performance was the same.
based on some reviews on Amazon, the value in the product vs pads is avoiding the need to find the correct thickness of the pads. it also doesn't cost much vs branded pads
i was hoping you would do vram specific hotspot measurements rather than FPS improvements, the 3090 struggles hard with keeping the vram cool on the backside of the gpu with just a passive backplate
I must say this stuff is amazing, for a while i had problems with my gpu of 5 years overheating and constantly on being 94 degrees. I already changed the thermal paste on it a few weeks back no improvements. Bought this stuff after the video and now on maximum loads it runs around 76 degrees max. Like i said amazing stuff and thanks for the video!
Was one tub enough for the GPU vrms?
@@Barracuda1807 more than enough for my gpu
@@zizoraf What did you use on the core? Same stuff or?
@@exterminater267 If i remember correctly i also put it on the core, and still working fine after a month
There is no way you got a 20* delta with just a different paste. Something else had to have changed. Otherwise you got god tier paste and would be a game changer. Just sayin.
I think a part 2 would be awesome where you took a card that famously got hot and see how good this stuff would do with it.
Love seeing LTT publish these videos. Can’t wait to see what’s to come with their new space.
I'd love to know the thermal data on this stuff.
I've changed the thermal pads on a few high end gpu's recently, and I'd be curious to find out if the k5 pro has any advantage over a good 12.8w/mk pad
From what others are saying, it will perform worse but has the huge benefit of not having to know the thickness of the pads, just apply a ton and you're done.
I have a silly question... Does it affect coil wine because it has more tight air gap between the coil's and the heatsink affecting the vibrations...its also like silicon based thing so it may isolate some noise...that would be interesting of using this thing for that reason
I'm hoping other reviewers will cover this. Very tempted to pick some up for re-pasting my water-blocked GPU. It gained a pretty nasty whine with only the EKWB pads in place.
No it would not affect coil while, coil whine develops when there is electronic magnetic field interference from your other components or can be triggered by the quality of the power delivery from your electricity to your rig. And can also be affected by the quality of your power supply too.
I recently upgraded my power supply (from a 850W to an 1000W) for my 3090 pc and it coil whining minimize dramatically. Maybe that helps!
I don't believe so just having a tighter frame would do the job
@@mikeycrackson in water blocks coil wine is often amplified by the big solid block compared to the stock air cooler. I have the issue on my 2080ti, which is significantly louder on water. I think it would be interesting to know if there is an improvement.
great video, just a thought could this overtime loose it's effectiveness? Or does this magically work forever? It would be interesting to see performance over time 6 month, year, two years.
Serious question.... Do you think this would work with the 2060 Ko Ultra? That thing has 7mm pads and is doubled in some areas.
Whats interresting is the thermals not the FPS.
Does it improve FPS by 6% under same thermals or what?
I’d drop 1fps anyday if i drop 5°c for it. I mean, more in depths testing would have been awesome.
I don’t understand how this would result in better performance unless 1) it was thermal throttling stock or 2) you were able to apply a greater OC thanks to the extra thermal headroom
That's what I was wondering.
I was wondering the same, should have pushed OC to its limits before tests to see any real difference. Should have just given the temp / fan speed to show sicne that really the only difference this can make
great info thx, btw is this paste usefull by combination with watercooler ?
This is really interesting although it doesn't seem to be available in the UK. I have a Dell XPS 15 9560 which is throttled massively by the VRMs overheating and since BIOS updates have nixed software control of voltage, I didn't have an alternative to using pads to try and take the heat away via the chassis. This looks like it would probably help. I already replaced the stock CPU and GPU paste with Thermal grizzly which made an appreciable difference to gaming temperatures.
I personally wonder more about the difference / benefits between a normal thermal compound and this product?
Probably worse. The goal is to be a better substitute for thermal pads, but is messy (which is why you'll likely never come across this in place of thermal pads).
This stuff has existed years; if it was better than normal thermal paste it would be more commonly known.
Interesting, totally not worth it but interesting. 👌
Totally worth it if you need to change the thermal pads. Just use this instead.
I don't think Ltt has made a video on a product worth it for me in years. Still watch daily lmao
Thermal pads are actually a pain in the butt though. If you don't have the correct thickness pads, or even know what they should be, this product would be a great thing to use. I'm also thinking it might do a good job of reducing coil whine, since it's designed to be applied heavily.
Indeed. Though he made a good point at the end. If you were planning a repaste anyway, might as well add some of this
Absolutely worth it.
My laptop fan is dying, to replace it I need to take out the thermal module, when taking out the thermal module you need to replace the thermal pads.
Those thermal pads need to be an exact match to the OEM thickness. How thick are those? Who fucking knows, the laptop manufacturer gives no specs, and as far as I can tell, there’s not even an easy way to purchase the OEM one so you can be sure they will be right.
So what are my options? Buy this stuff like I did, or I could buy a kit of every single thickness of thermal pad available just so I could be sure I will have the right one, while hoping that I actually managed to correctly figure out the thickness when we are talking about submillimeter differences.
Or I could take it apart, try to measure it, hope I get it right, then wait around with my disassembled laptop until the new pads arrive and again hope I got it correct?
Even if there’s no performance gain whatsoever, the convenience alone makes it worthwhile.
i did this on my laptop almost a year ago and when i used it alongside kryonaut v2 it brought my system from thermal throttle and raging fans down to quiet fans and maintaining full boost speed.
Can this substitute both .1 thermal pads and .2 thermal pads (for water blocks that require both) at the same time?
The timing on this is amazing. Literally loading up a cart with liquid metal and looking at different thermal pads with plans to repaste and repad several laptops. I think I will go with this stuff instead of the assortment of pads.
@@orilion1820 That would be an interesting insight if I ever used the feed. I only view my subscriptions page. I saw this video posted while shopping and searching for something to throw up on the second screen.
@@orilion1820 Uncommon for sure. I doubt my purchase preferences are directing LTT upload schedules though. I know well the nature of data collection across different services resulting in suggestions. There are virtually no coincidences in recommended feeds. Whole reason I use the subs page is the direct serving of content in reverse chronological order. No need for a recommendations feed when friends and content creators do plenty of that already.
Hmm, depending on how thick it is, you might be able to use a piping-bag for getting it on there.
It will probably reduce the mess, and application-time, quite a bit.
I was thinking the same thing kind of. Using different tip size syringes to apply it
I've always just used non-electrically conductive thermal grease. You can buy a large tube of it for cheap.
Is there any problem with this?
Maybe I just noticed, but the lamp on the desk looks like outer shell of an early iMac!! Soooo cooool!!!
yes you can see the creation in the vid office makeover mkhb or smth they did each others offices.
This seems really great honestly - assuming that it does indeed stay stable and retain it's shape over time. That would be my only worry since it's a "goop".
Better thermal contact is of course good, but I think the real big benefit here is the versatility to bridge any size gap. You never have that one pad thinkness that you need...
It is not suitable for gaps wider than 1mm for long term use. Over 90°c airpockets form.
@@HG-fw8hn What is your source on that? Experience? It's advertised as up to 3mm.
@@TheStigma Personal experience. VRM's in a 100w+ RTX 2080 laptop. 2mm clearance. Applied more than enough, ran for 2 weeks then black screen and no start. Removed heatsink, nothing but airpockets in goo and burnt out VRM's. 550 USD repair. Stuff is no good for gaps over 1mm. Should only be used in PSU heatsinks and industrial heat exchangers in my opinion, fine for that application.
I'm using this for a couple of years now in my 2011 iMac with zero issues. You can upgrade the MXM GPU in these, but the heatsink obviously not designed for this and its pretty hard to find the correct size pads because of the weird thicknesses MXM GPUs have and this was super easy to apply. The GPU in these machines sits vertical and it retains its form/shape just fine, it doesn't flow out or anything. You can even scoop it off the old one and reapply to the new GPU if needed, it has a really weird but fun consistency.
Problem comes than you dont know the thickness in loptop as its nowhere written in manuals only in service manuals which they wont give you. I have very old loptop its termopaste just cracked and fall off after i once checked and opened it. You end up with old loptop which crashes or wont even boot or boot bios if you put wrong stuff and to find out even the thikness answers in world wide web will be random people guessing and saying you must add this or this. Next problem you need to order it.
I'd really wish they mentioned the average temperature difference of the memory of the GPU with the pads and with the K5 compound. See if there was a difference there. Also thermal paste temps too with the core.
This is exactly what I was looking for as well. My 3090 memory temps get really high and knowing if those could be improved would have been really useful.
I would love to see if mem temps was reduced to. I have a rx5700xt that has high mem temps, thinking of testing this stuff out on it..
@@jzetterman I'll spare you my pain, they make memory temps worse, my 3080 has gone up over 10 degrees, it doesn't work
@@Alpejohn I've got reference 5700xt and I'm curious too
K5 pro is very underrated. It solved VRM overheating in my modified Clevo, nothing else seemed to work. Hope it gains more traction. Thanks for the video on this (so far) niche product!
I recently bought this not long after seeing this video and finally put it on my card the other day. After some trail and error adding more because I didn't realize how little I had on there apparently not only is my GTX 970 card running stable with a +250 core overclock but runs over 10C cooler even with the fan speed lowered to reduce noise. I am very happy to have learned about this and my video card thanks you too. Now I can finally play games without a turbine spinning up in my case and can get rid of these cheap, crappy thermal pads that make it so I could barely overclock it at all after ripping the stock ones. Also yes I'm running a 970 and still happily gaming since I don't really care for most of these newer titles or gaming at 4K. My stress test by the way was about half an hour of GTA Online just driving through the city and three hours of VR and they both ran noticeably smoother then what I had been getting.
I am from Greece and I could never imagine myself seeing a product produced in my hometown (Thessaloniki), to be showcased in a Linus Tech Tips video!
Was thinking the same thing! I’m from Kavala area though
@@amoiridis Εκλαψα οταν ειδα το πλανο της ΕΡΤ. Σαν τον hardware canucks που ειχε ερθει θεσσαλονικη και ανεβαζε φωτογραφιες.
“So what’s in your laptop”
“Oh normal tech stuff, like toothpaste”
You didnt show how it looks when you take the heatsink off after the K5, do you need to re-apply like thermal paste? does it solidify a bit in the shape of the components?
You guys need to do a follow up on this... I recently did this on my 3080 laptop and it was great for a few days... Then it was crashing a lot. I opened it up again and found that the K5 pro for the vram was all stuck to the heat sink and boiled out to bare metal in spots. It's like it separated from the vram and hardened on the surface. I temporarily switched it to thermal paste and the crashing went away. Now I'm left with a mess and waiting for thermal pads to come in the mail.
Oh that's really interesting! That stuff sits on my desk for more than 3 years. I've only used it on an old gf8800gts so far, but it seems to work great, especially on BGA RAM and around 1-1.5mm, maybe 2mm gaps
would love to see this in a later full waterblock setup, and also comparing on CPU VRMs on boards that have super chonk VRM heat sinks
So does this particular paste replace any thermal solution? Or is it in addition to? Example, if im repasting a GPU, do i add this directly to the die itself? And do i add it to the pads? Or does this take the place of the thermal pads and traditional thermal paste?
Linus only give you a 15 minute break for lunch? What a ball buster!!
unusually, this is exactly the kind of thing I've been looking for for the better part of a year.
I was in the middle of refurbishing a Powerbook g4 17 inch that my grandma gave me. It's an ancient laptop, and uses some kind of white thermal goop instead of thermal paste or pads to cool the GPU. I was never able to find the right size pads or paste that was similar to what came from the factory, but this looks like it might do the job.
white thermal paste? Sounds like nano diamond or IC diamond if I remember right. Standard thermal paste in the old days compared to the gray Chinese goop nowadays.
Could also be thermal epoxy.
In any case, you shouldn't need thermal pads for the VRAM and any Chinese goop should even work provided your heatsink heat pipes still are okay and the fan still pushes air properly. You may also want to check on mounting pressure.
For something that old, I'd also look into oiling the bearings of the exhaust fan.
im liking alex solo-hosting these engineering videos! he makes the content interesting and accessible to an audience with no engineering background and i like his pacing and general approach :)
It makes Linus look like a really good boss. He assigns tasks, delegates and allows his employees to shine instead of trying to make everything about himself all the time.
I also like Anthony.
Is this stuff electrically conductive at all, or can it be liberally applied to a desktop CPU without concern for spillover?
First off love you guys. Secondly I would recommend heading to your local pharmacy or Walmart and getting an empty medicine syringe to apply this. Hope you guys have a wonderful new years and hope the next year bring us all some exciting new or old tech.
I tried this one my Gigabyte Gaming OC 3090, on the RAM. It was basically the same as stock and still thermal throttled. I ended up using Thermalright Odyssey Thermal Pads, 2mm thick ones. Those actually dropped the temps by like 12 degrees and made the Ram on my 3090 stop throttling. Now this is probably better than some of crappy thermal pads you get with stuff but it definitely doesn't beat a high end thermal pad like the Odyssey or Fujipoly pads.
I like this video overall, but a point I think that was missed was discussion or mention of thermal conductivity. According to K5 Pro's Amazon listing, the paste has a thermal conductivity of >5.3W/mK whatever "greater than" that value actually means. Given this number (5.3), we'd expect worse heat transfer on anything with a lower value, so if you instead replace the stock pads with higher rated pads (like from ThermalRight or Gelid) which reportedly have values of 8 W/mK and even 12 W/mK, we'd expect better results than the KD5 Pro paste (assuming the heatsink is still capable of dispersing that heat).
So what I'd like to see is a follow up to this where the KD5 is replaced with those higher thermally conductive pads.
point is thats hard to get the info which thermal pad thiccness is the right one for your device, so the k5 goo is just the better alternative than aftermarket pads. Wrong thermal pad and the device can literally die
Look into graphene pads and how thermal paste is still better
Its not all about that number necessarily the contact area has a lot to do with it as well which isn't as easily quantifiable but having something with the ability to flow and squish together will generally have better contact than a pad that wont. Its the reason that thermal paste is so effective
@@connorjohnson4402 exactly. Graphene pads have a thermal conductivity 5-20x that of traditional paste, and still loses to said paste. Contact makes a bigger difference than almost anything else. If contact is exactly the same, then you can compare thermal conductivity. This is why comparing the thermal conductivity of two different brands is also bad, as they have different chemical makeups.
@@faranocks Don't graphene pads also tend to have better thermal conductivity laterally than vertically?
can we use this on i5 11400 stock CPU stock cooler with stock paste? my chrome is shooting 56. idle stays 30.
I'm from Greece and I actually never heard of it. I'm gonna buy and test it now. Thanks ltt 🤘
doesn't the 3080 and 3090 have problems with thermal pads anyways I've heard people say they have a performance upgrade by using different thermal pads I guess the stock ones just suck
I know they aren't very good on the founders editions
I've always wondered if putting thermal paste on thermal pads would be effective.
That's like putting 2 condoms. Sometimes less is more.
if you think it doesnt make good contact you can absolutely do it to fill the gaps
i did it on my 3080 and it dropped a few degrees on the mem junction
The only reason pads are used on the memory is clearance. Paste might not fill those gaps adequately.
It can be depending on circumstances really, I have done that on pads I had to reuse at several occassions. But it was always just minimal amount of paste there. Also results did vary, had cases where it helped driving temps down and had cases where temps went up, then needed to go back and clean it... pain in the ass. But in general, I only did it when reusing old pads, which were already deformed by long time spent on VRM's or memory chips.
It's quite possible that my varied results also depended on different pads material/quality/thermal paste used.
So I'd say to only try it if you want to monitor temp change and don't mind cleaning it if it mess up your cooling.
@@Tubenstein yes i also used one tiny drop of paste per gddr6 chip on top of the used pads and it worked very well
Would they get better performance if it's been used for a while?
Why are there lines slowly moving up the background? Are the lights flickering around the shutter speed or frame rate???
This stuff helped loads with my ASUS GX502GW. Worth noting that the manufacturer used a similar "thermal goop".
Ran furmark for 17 mins and it was consistent throughout.
CPU still hits 90 and above on really demanding work loads but it's more of a brief spike that dissipates faster.
Better than using pads on the VRMs and other chips which I originally used when refurbing my laptop as I didn't know about K5 at the time.
I used thermal grizzly paste on the CPU and GPU.
the first 30 seconds of your video taught me that if you are able to keep the memory modules cooler than the rest of the board you will get better performance...
THANKS ALEX i will purchase some of that stuff for when i rebuild my R9 Fury Nitro
not sure if i missed a part of the video but have you tried other mainstream thermal pastes (kryonaut etc) as thermal pad replacements to compare with this k5 stuff? kinda curious if theres any difference
WARNING: Before you buy K5-pro and put it all over your GPU like I did, reconsider if you have DDR6X VRAM (3080/3090 variants). DDR6X runs too hot for the K5-pro to properly conduct the temperatures when you are doing anything memory-intensive (mining, rendering, etc.). The memory junction on my 3080 was reading 110c and causing the card to thermal throttle. I ordered 15W Gelid ultimate thermal pads to put on the memory units. I am not looking forward to removing the K5-pro paste. I’m going to leave it everywhere else since 5.3W thermal conductivity seems good enough for the other components.
How's the Gelid Ultimate? Is it good?
How'd the Glid go? Didn't work for me but idk, tried 1mm ram was at 107 degrees, checked and there was a chip with bad cover, tried 1.5mm and the same temps... Did take off the plastic both sides
You may want to look into Penchem TH949-1, Upsiren UX Pro, CX-H1300, or Laird T-Putty 607. Very nice putties, and great for cooling GDDR6X
Seems like a good product.
Replacing thermal pads kinda sucks. You need the right thickness, then need to cut it to size, etc.. Just putting this stuff on seems easier.
The replacement ones are also pretty expensive, and you need a pack for each thickness. Plus all the time spent to figure out what thickness you need.
I'll definitely try this stuff out when I do my overhaul this summer
@@KPalmTheWise I think thermal pads would be a bit cheaper. You really only need 0.5mm, 1mm, 1.5mm and 2mm.
Sure, you can get a giant tub of this, for USD150 but you probably wouldn't want that much, so you'd have to buy smaller amounts.
Yes, there is 3mm, 4mm and even 5mm, but this stuff likely isn't going to work for that anyway. As he said, at that point you are going to use copper shims(which you'll have to buy as well). So I don't think that is quite comparable anymore.
Still, I stand by my opinion, but I don't think you'd save actual money with it.
How does this product stack up against Laird’s thermal putty? The miners have been using this for quite some time when servicing GPUs, since some manufacturers use thermal putty instead of thermal pads (because they can’t reach desirable tolerances on component heights/ cooler manufacturing).
Thank you for showing me this, I just bought a ton of it and I'm using for every thing that uses thermal pads and WOW what a difference.