I really wish the AIBs certified thermal pads and thermal pastes/putties for use in reconditioning GPUs so there wouldn't be so much guesswork , which if it turned out wrong in terms of thickness and stiffness could break the GPU , or the VRAM from the board and turn hundreds of dollars into a paperweight. Would be so much easier if you bought some thermal grizzly and it was "approved for EVGA FTW3 models" or something like that
My stock PowerColor 6750XT Fighter had pads that were too thick from factory which is a bummer for the GPU Hotspot temps. Cooling solutions for GPUs definitely should be as expansive and easy to understand as those for CPUs
GELID thermal pads are insanely good value; best for its price range. TekShinji did a comprehensive benchmark of several after-market pads, and the Gelid pads topped the chart in price/perf. But unintended effect of good pads: the heat still gets dumped onto the same heatsink -- so while it will help lower your VRM temps, it has an adverse effect of raising your core GPU temps at equilibrium a few degrees higher. Not too important in the grand scheme of things, but something to be mindful of.
That´ll heavily depend on the cooler of the GPU. On something like a TUF 3080 were the memory and Core don´t have a directly connected cooler it improves both, as the thermal paste you´d use is likely better too. Also On my Strix card I´ve only seen lower temps on the core since I´ve replaced the pads. There is a reason for that too. Thermal transfer on a copper cooler is way higher while it has a lower temperature. Meaning if there isn´t 90-105°C memory but 68-74°C memory (numbers from my personal experience) then the cooler does have a lower overall temperature, meaning the core temp isn´t actually negatively effected at all.
polished copper shims on vram and the gpu die actually significantly drop temps over pads they need to make a pad with copper in it if they can make copper stretchy wastebands then a copper thermal pad should be no problem
I replaced my thermal pads with copper shims, it's much better than any thermal pads I ever buy. The temperature drops from 80 C to 65 C. hotspot from 95 C to 75 C and VRAM temp from 82 C to 70 C compared to good thermal pad (Thermaltake). Copper has much better thermal conductivity and much cheaper too. I bought 10 pieces for less than 7$ total. The installation however is much harder than thermal pads. It's not for everyone.
Just make sure you measure your thermal pads before you order. Too thick of a thermal pad will cause more issues than not changing them. Always check with your own card model and model number as each cooler and each 3000 series card will have a varied amount of thermal pad thickness. The aorus master 3070 to has 3 sizes, 1mm, 1.5mm and 3mm (for the backplate)
Gelid pads are great!! Reduces 10c to 20c on air cooled cards VS stock pads depending on the make. Gigabyte pads aren't great on GDDR6X for instance on their Eagle models
I changed the pads on my old 1080ti - bought it used in 2019. Dropped temps on mem by 4c.... not amazing results for the dollar... but fun to try out :)
Gelid Ultimate 15W/mK pads are awesome. I've watched many comparisons between pads and Gelids were consistently the best, or within 1C difference with the best. While being much cheaper than the other premium thermal pads. I'm using those pads currently on the backside of my GPU after installing a big finned heatsink on the back of my 1060 6GB. It does its job really well. The heatsink on the back gets really hot to the touch without airflow (can't touch it for more than a couple seconds because it's hot). It definitely helps with cooling both the memory chips, as well as the GPU core. The effect would be even greater on more powerful cards than my 1060. I'd recommend doing this mod to anyone really.
I’ve switch my stock thermal pads on my rtx3080 and Ti models and seen between 7-12c lower temps under gaming loads. My Ti had some “meh” stock pads which actually replaced with Gelids and saw about 11-12c drop in memory temps. My Ti isn’t water cooled and using its stock 3-fan cooler. My 3080 is running about 7-10c cooler under gaming load and slightly lower again once I placed on an EK waterblock 😇🥰👍
@Swim Fan Yeah Linux is great, i use now Pop OS on both my computers, Game-PC and Mini Media/Console-PC, and i do not want to go back to Windows 11, it just works that great, and games run fantastic.
@Swim Fan Strange i do not have problems that constantly extensions keeps breaking or not working correctly, if that was happening, i would have not use Pop OS anymore. Yeah what do you expect from Google (Alphabet Inc), they life from stealing our data, and much more.
Just repasted my 2080 ti and didn't touch the pads I just put new paste on the GPU and put it back togeather. Temps went from 82 down to 67-70 under full load. NICE! With that said the GPU came apart real clean and the old pads were completely fine, if they had ripped apart I would have changes them also but...they didn't :) All in all a great results with the old pads!
If you have a metal back plate, it help to put pads on the back side of the PCB to move some additional heat from the other side of those memory dies. I don't remember my exact results, but it was enough that I was happy with the time/effort invested. I also put some copper heatsinks on the outside of the back plate to help more. That GDDR6x makes some heat.
Did the same thing with Gelid2.0mm on a used 3080 FE bought off eBay. It reduced the hotspot by 19 degrees. I don’t even worry about the heat anymore. Also, the FE cards are the worst to open up.
I got my 6600XT *almost* new. The card had only been on the market for a couple months at the time, I think the guy wanted to scalp it but found he couldn't get a crazy offer. Doesn't matter how new your card seems, though. It's always worth checking! Pads like these are cheap enough and are worth the peace of mind they offer. As well as being handy to have in your own personal Tech YES repair kit.
It's important to measure the results under real-world use cases, not just synthetic memory tests. In many cases having much better thermals on the memory can actually lead to higher core temperature because the cooling solution is having to absorb much more heat. My 3080 Ti FE with new pads did drop memory temps down a staggering 20-30 degrees, but it also raised core temps by about 5-10, and this is with the exact correct thickness measured by myself and very careful and precise mounting of the cooling solution. This may not matter if your core usually doesn't get hot, but if you play at high resolutions and push the card you could actually be losing performance due to reduced thermal headroom on the core. If you are going to go the distance with new pads it may be prudent to go all the way and do a water cooling solution that can dump all that extra heat. I sort of regret my 3080 Ti FE. The FE cooler just does not keep up, and the stock pads are embarrassing.
In recent years I've replaced all GPU and MB pads (VRM, chipset/PCH, VRAM) with GP Ultimate in both my last two machines. Even without sensors, using both the 'finger test' and an IR thermometer it's quite clear that (in most cases) there is pretty massive difference - the sinks are much warmer, 10°C or more. Unsurprisingly it's most pronounced where the original pads were the cheap, rubbery silicone ones. Gelid GP Extreme and Ultimate are effectively pre-formed strips of thermal putty rather than 'pads' as it's usually understood, and if you check out the price of good thermal putty they're very good value (and certainly a lot more convenient and neater than putty).
I've seen from several sources that the cheaper GP-Extreme can actually be better. They are a softer material so they conform better and aren't as strict on exact pad height. Just a thought.
VERY IMPORTANT !! The thickness of the thermal pads are can ,and WILL BE different. In order to know it,one of the sources is the tech support of the card maker.
It depends on the card, but I changed both thermal pads and thermal paste on my old 3060. Temperatures were through the roof beforehand because of TERRIBLE factory pads and paste from GIGABYTE. I only changed paste on my 3070 because the factory pads from MSI were decent enough at keeping temps low. But the paste was obviously not up to my standard. :)
i changed the ones in my strix 3080 for a copper plate that was machined specifically for it. it dropped 30C easily. was at 105C and now its between 65-70C
Gelid makes good stuff, check out their mini fan header to standard fan header adaptors, if you need to run a fan off a GPU PCB for a Hybrid setup so it's all self contained.
I spent some time looking into this and the issue is the 2mm thickness of the pads in the reference design. My gigabyte model has a heatsink with a raised heatsink so the thermal pad can be 1mm. Copper shims from aliexpress with paste between the shim and heatsink and thinner pads would give better thermal transfer.
Great video Brian. I've watched a few videos where they replace these pads with copper shims, and reduce temps around 20c or more on the 3080 & 3090 cards. Would be good to see you give that a try for comparison.
What temperature are you getting? I was getting about 70c while i was mining with the card which seemed fine to me. Kinda lazy to open the card but the thermal paste is due as well it's been 2 years since i bought it.
Even at 64 degrees, that's a long way off throttle temp for the G6X. I've always thought G6X ran really hot, but maybe it's just this cooler is really good or they used better pads? But if you're buying a used 3070 Ti on up... I might not consider it if they haven't done something for the memory since it's really easy to hit throttle temp when mining. I know there are a lot of better options, all the way up to copper shim plates that really help with cooling.
I did a research as well back in the day about this like half a year ago... but I wasn't sitting on any High End card, I was using MSI Gaming X RX480 8GB for a quite some time, and I was pretty concerned about the thermal pads after changing my thermal paste, so I looked for something to replace them, just for the sake of getting peace of mind... and I did order the same pads that you did there but different thickness to match the originals. After replacement my Memory Temp was slightly less... and core also dropped few degrees, so it didn't change much, but I think it was worth it anyway... I replaced my RX480 for RX6600 on my Main PC though, but I think I will do the same course of action with pads after some time anyway...
I finally got around to changing my 3090fe out a couple days ago. I used the thermalrights (3 pads) 1.5mm and went from 104c to 80c. Can't believe how expensive this card is and had to do this to keep from nuking my system and prevent thermal throttling... Ridiculous... 😮💨
hopefully you could score some Aluminum Nitride Pads and test it out, I've read some articles online that it has the highest w/mk compared to the rest of the thermal pads being sold online
I'm really more curious how the stock ones compare to some really cheap 1 dollar pads, I've used plenty of those, and they always seem to do okay, but I have never properly tested them, I might give it a try on my 3080 since I have quite a few left
MSI are easily the most conscientious when it comes to the pads on their cards - try a pad swap on a Gigabyte or EVGA and the difference is much more marked.
I thermal pad modded my 3080 FE with 2mm Gelid Extreme pads and saw some improvement. Two weeks later, I copper modded with 1.2mm plates ($10) and thermal paste on either side and the improvement was much more significant. I don’t mine and cap my frame rate to 60fps while gaming. I just want my GPU to run cool and last a long time. My next rig will be in at least 5 years.
I got my brand new msi 3080 gaming trio 10gb. Came not new, not sealed. Definetly repackaged. I dare not return it though since i won it on a shuffle. After watching a tear down video i notice my cards missing thermal pads all over. Would love to repad and repaste it.
5:08 I would not change the tape in this condition. And it seems to me that the tape that nVidia uses is much better, so two reasons I wouldn't go into this...
Be curious to see the same test done but with the copper shims mod - use copper sheet the same dimensions as the thermal pads with the bare minimum of thermal paste applied between the memory and copper shim and then also between the copper shim and cooler. I’ve seen this done in other vids but been dubious with the results - they seemed to good to be true.
Should've added kingpin paste on memory, then place the pad, then add more paste to finish the sandwich will give you better thermal transfer. Also, the gelid pads are too hard, so look for softer pads like Valor Odin.
I’ve been checking eBay here in Oz for used 3080 and 3080ti’s. At the moment there’s only a couple for sale. One local guy wants the same price for his mined card as a new one. I offered him aud$1000 and got no reply. I’m not sure we will see many mined cards on the market as these guys over invested when the prices were high and can’t come to part with the cards at a realistic second hand price.
@@ChrisZukowski88 yeah i could care less thats why i offered him 1000. pretty generous for a second hand mining card i thought? He will probably end up keeping it, I would espacially if i paid like 2500 for it lol
I've replaced mine on my 2070 with these, they're great it also looked like my specific card had some missing on a set of vram. I have also heard after a certain point you might have deminishing returns as the backplate cannot dissipate the heat quickly enough. Would recommed 10/10 mine dropped about 10C
I had great success in that my VRM's didnt have pads so I added pads onto the VRM's and my GPU hotspot temperature decreased by about 10 degrees at full load. I would have thought that we are loading the cooler with more heat so that would increase the GPU temp but actually no, for some reason, the memory and gpu temps were great. oh, I have a 6700XT Mech X2 OC and was having issues with driver timeouts before that. I am thinking my next step may be to make the backplate a passive cooler by putting pads inbetween cos at the moment it is just for show.
The VRM is thermally well coupled to everything on the GPU with a lot of the coupling area having little opportunity to remove the heat, it's essentially forced out through the components. The coupling occurs via the massive ground and voltage plane pours in the inner layers of the PCB. Heatsink actually usually has a fair bit of thermal impedance between the different areas and it can absorb that extra heat no problem either.
So, forgive me, I'm but a newbie at these things. But you compared a GPU with (possibly) dried out (and certainly not new) thermal paste with the same one with new thermal paste and pads? Is this correct? If that is the case, how do we know the effect of the pads?
The original pads on your 3080 look good anyhow and you are water cooled?. The FE 3080 had some nasty oily pads, and mem temps would easily hit 100C, especially in mining. That's why most people did the pad swap. For myself and my original FE 3080 swapping the pads made a huge diff to the VRAM (like over 20C). Incidentally GPU temps don't drop and could even go up a little as you are now dumping more heat into the heatsink. You do need to make sure you get the right pad size as it can affect the GPU die mounting pressure. I (and others I know) have had good results with Thermalright pads.
Could you revisit this and add thermal pads on the back so that the backside of the memory contacts the backplate? And see how much difference it makes since gddr6 is flip chip and the hot side is at the back
if the back plate was not paded its going to be tricky as he will have to figure out the thinkness of the pads i know from seeing miners change pads yes changing the pads on the back plate does help too even on my 3060 that i redid it helped
Fyi to people out there if you've bought gigabyte graphics cards this is definitely something that should be done as stock thermal pads leak oil and are known to cause issues or loose cooling performance.
hey tech city which is better extreme vs ultimate pads wanna buy some and do you know which thickness for rtx2070 super fe and rx5700 xt also a fe model
I replaced my stock ps4 pads with fujipoly ultra extremes and while I haven't measured, it does seem to have helped transfer more heat to the mid-plate. I compensated by cutting some holes and adding some fans. Looks pretty hc but it I like it!
That sounds more like a GPU problem if it hits high temperatures it should shut down it shouldn’t die. Unless the card it was already faulty. There’s a few good UA-cam videos that show you all these different brands the squishiness and thickness of each so you can determine the correct thickness. Some of the 2MM are squishier than others or Stiffer.
If you mod something, you have to be prepared to buy it twice if it breaks, which begs the question, why didn't you just buy the better model, to begin with? People who mod are either rich, YOLOers, have had experience with other cheaper/expendable parts, or are UA-camrs with warehouses full of parts to work on.
@@eliashowe7115 dude many GPUs didn't have good paste or pads, even on top of the line GPUs. Not to mention that weather + humidity condition during shipping / storage can also affect their quality.
Can you test Gelid Ultimate vs Thermalgrizzly Minus Pad Extreme? Those TG pads are an upgraded version of their old Minus Pad 8, and they are ridiculoudly expensive. I think 3x the price of Gelids. I can't find anyone who's willing to test them. My GPU doesn't have thermal sensors on memories so I can't really do it myself.
Can you help me for the Gelid GP-Ultimate thermal pad thickeness for VRAM, Backplate and those weird row of blacks from my Gigabyte RX 5700XT 8GB OC? the stock pads were dry and not sticky anymore.
For a top mounted radiator (for exhaust), do you recommend a push or pull configuration? Currently it’s in a push configuration but it’s damn noisy and wish to quiet it down a bit
This is very dependent on the case you are using/ config of the other fans/radiator placement. Need more info for very accurate advice. In general tho positive pressure in the case is best
I don't think that the test was accurate ... cuz on the first one the thermal paste was dried and on the second one it was fresh. which could be the only reason for the temp change. correct me if I'm wrong
I really wish the AIBs certified thermal pads and thermal pastes/putties for use in reconditioning GPUs so there wouldn't be so much guesswork , which if it turned out wrong in terms of thickness and stiffness could break the GPU , or the VRAM from the board and turn hundreds of dollars into a paperweight.
Would be so much easier if you bought some thermal grizzly and it was "approved for EVGA FTW3 models" or something like that
Very good idea, email thermal grizzly.
My stock PowerColor 6750XT Fighter had pads that were too thick from factory which is a bummer for the GPU Hotspot temps. Cooling solutions for GPUs definitely should be as expansive and easy to understand as those for CPUs
I changed mine because I ripped them off by mistake.
Pads are probably $10 for like a3 size hahaha
@@RYU47376 Depends if you are paying more for higher thermal conductivity.
If the ones that tore look and feel like tofu, then it was no mistake.
If it's from ali express it's most likely going to be fake.
I love these kind of videos. Not enough youtuber cover after market performance improvements.
GELID thermal pads are insanely good value; best for its price range. TekShinji did a comprehensive benchmark of several after-market pads, and the Gelid pads topped the chart in price/perf.
But unintended effect of good pads: the heat still gets dumped onto the same heatsink -- so while it will help lower your VRM temps, it has an adverse effect of raising your core GPU temps at equilibrium a few degrees higher. Not too important in the grand scheme of things, but something to be mindful of.
That´ll heavily depend on the cooler of the GPU.
On something like a TUF 3080 were the memory and Core don´t have a directly connected cooler it improves both, as the thermal paste you´d use is likely better too.
Also On my Strix card I´ve only seen lower temps on the core since I´ve replaced the pads.
There is a reason for that too.
Thermal transfer on a copper cooler is way higher while it has a lower temperature. Meaning if there isn´t 90-105°C memory but 68-74°C memory (numbers from my personal experience) then the cooler does have a lower overall temperature, meaning the core temp isn´t actually negatively effected at all.
core being 3*c higher won't affect lifespan, vrms and memory are much more sensitive to heat
polished copper shims on vram and the gpu die actually significantly drop temps over pads they need to make a pad with copper in it if they can make copper stretchy wastebands then a copper thermal pad should be no problem
I don't see how this is an issue considering that's the entire point of those thermal pads even being there.
1:14 It's shaping up to be EXCITING times for Tech YES parts hunts with the flood of GPUs incoming!
I remember the "70$ rx470", so i can't wait for the 170$ rx5700 haha
I did this on my old rx 580 and it worked extremely well. Bought 17w\mk pads and they worked great. Drastically brought temps down.
I replaced my thermal pads with copper shims, it's much better than any thermal pads I ever buy. The temperature drops from 80 C to 65 C. hotspot from 95 C to 75 C and VRAM temp from 82 C to 70 C compared to good thermal pad (Thermaltake). Copper has much better thermal conductivity and much cheaper too. I bought 10 pieces for less than 7$ total. The installation however is much harder than thermal pads. It's not for everyone.
Just make sure you measure your thermal pads before you order. Too thick of a thermal pad will cause more issues than not changing them. Always check with your own card model and model number as each cooler and each 3000 series card will have a varied amount of thermal pad thickness. The aorus master 3070 to has 3 sizes, 1mm, 1.5mm and 3mm (for the backplate)
Gelid pads are great!! Reduces 10c to 20c on air cooled cards VS stock pads depending on the make. Gigabyte pads aren't great on GDDR6X for instance on their Eagle models
This is the video I need as I upgrade from my 1st owned 3070 to used 3080
I changed the pads on my old 1080ti - bought it used in 2019. Dropped temps on mem by 4c.... not amazing results for the dollar... but fun to try out :)
Gelid Ultimate 15W/mK pads are awesome. I've watched many comparisons between pads and Gelids were consistently the best, or within 1C difference with the best. While being much cheaper than the other premium thermal pads.
I'm using those pads currently on the backside of my GPU after installing a big finned heatsink on the back of my 1060 6GB. It does its job really well. The heatsink on the back gets really hot to the touch without airflow (can't touch it for more than a couple seconds because it's hot). It definitely helps with cooling both the memory chips, as well as the GPU core. The effect would be even greater on more powerful cards than my 1060. I'd recommend doing this mod to anyone really.
I’ve switch my stock thermal pads on my rtx3080 and Ti models and seen between 7-12c lower temps under gaming loads. My Ti had some “meh” stock pads which actually replaced with Gelids and saw about 11-12c drop in memory temps. My Ti isn’t water cooled and using its stock 3-fan cooler. My 3080 is running about 7-10c cooler under gaming load and slightly lower again once I placed on an EK waterblock 😇🥰👍
@Swim Fan Yeah Linux is great, i use now Pop OS on both my computers, Game-PC and Mini Media/Console-PC,
and i do not want to go back to Windows 11, it just works that great, and games run fantastic.
@Swim Fan Strange i do not have problems that constantly extensions keeps breaking or not working correctly, if that was happening, i would have not use Pop OS anymore.
Yeah what do you expect from Google (Alphabet Inc), they life from stealing our data, and much more.
Just repasted my 2080 ti and didn't touch the pads I just put new paste on the GPU and put it back togeather. Temps went from 82 down to 67-70 under full load. NICE! With that said the GPU came apart real clean and the old pads were completely fine, if they had ripped apart I would have changes them also but...they didn't :) All in all a great results with the old pads!
Great video as usual! ;)
If you have a metal back plate, it help to put pads on the back side of the PCB to move some additional heat from the other side of those memory dies. I don't remember my exact results, but it was enough that I was happy with the time/effort invested. I also put some copper heatsinks on the outside of the back plate to help more. That GDDR6x makes some heat.
you uploaded this while I was replacing my 3090 Suprim X's pads with K5 Pro (thermal pad replacement paste)... wavelengths
Did the same thing with Gelid2.0mm on a used 3080 FE bought off eBay. It reduced the hotspot by 19 degrees. I don’t even worry about the heat anymore. Also, the FE cards are the worst to open up.
you look great bryan! after you left aus lol, no seriously your skin tone is shining
mine went from 95-105C to 70-76C switching from the original pads to these Gelid ones so it's definitely worth it.
Critical thermal pads are the best and are shipped in the exact sizes you need for whatever specific card. Dropped my vram from 105 max to 75 max
I got my 6600XT *almost* new. The card had only been on the market for a couple months at the time, I think the guy wanted to scalp it but found he couldn't get a crazy offer. Doesn't matter how new your card seems, though. It's always worth checking! Pads like these are cheap enough and are worth the peace of mind they offer. As well as being handy to have in your own personal Tech YES repair kit.
It's important to measure the results under real-world use cases, not just synthetic memory tests. In many cases having much better thermals on the memory can actually lead to higher core temperature because the cooling solution is having to absorb much more heat. My 3080 Ti FE with new pads did drop memory temps down a staggering 20-30 degrees, but it also raised core temps by about 5-10, and this is with the exact correct thickness measured by myself and very careful and precise mounting of the cooling solution. This may not matter if your core usually doesn't get hot, but if you play at high resolutions and push the card you could actually be losing performance due to reduced thermal headroom on the core. If you are going to go the distance with new pads it may be prudent to go all the way and do a water cooling solution that can dump all that extra heat. I sort of regret my 3080 Ti FE. The FE cooler just does not keep up, and the stock pads are embarrassing.
Thermalright makes good pads also. Used them in variety of components and the results always were great.
In recent years I've replaced all GPU and MB pads (VRM, chipset/PCH, VRAM) with GP Ultimate in both my last two machines. Even without sensors, using both the 'finger test' and an IR thermometer it's quite clear that (in most cases) there is pretty massive difference - the sinks are much warmer, 10°C or more. Unsurprisingly it's most pronounced where the original pads were the cheap, rubbery silicone ones.
Gelid GP Extreme and Ultimate are effectively pre-formed strips of thermal putty rather than 'pads' as it's usually understood, and if you check out the price of good thermal putty they're very good value (and certainly a lot more convenient and neater than putty).
It's a must if you have a high TDP Card (GDDR6X) and using it for mining, more of a recommendation for gaming.
I've seen from several sources that the cheaper GP-Extreme can actually be better. They are a softer material so they conform better and aren't as strict on exact pad height. Just a thought.
I did the same on my 3080 FTW because here in Mexico is a must since I got 100° C on stock pads and 92° C on gelid
VERY IMPORTANT !!
The thickness of the thermal pads are can ,and WILL BE different.
In order to know it,one of the sources is the tech support of the card maker.
It depends on the card, but I changed both thermal pads and thermal paste on my old 3060. Temperatures were through the roof beforehand because of TERRIBLE factory pads and paste from GIGABYTE. I only changed paste on my 3070 because the factory pads from MSI were decent enough at keeping temps low. But the paste was obviously not up to my standard. :)
i changed the ones in my strix 3080 for a copper plate that was machined specifically for it. it dropped 30C easily. was at 105C and now its between 65-70C
Got the specs for that? It sounds like a good mod
Gelid makes good stuff, check out their mini fan header to standard fan header adaptors, if you need to run a fan off a GPU PCB for a Hybrid setup so it's all self contained.
I've done this with laptops too. I have always used Fujipoly brand thermal pads but, yeah.
You should check out the K5-PRO viscous thermal paste for thermal pad replacement - it would be great to see a comparison
I spent some time looking into this and the issue is the 2mm thickness of the pads in the reference design. My gigabyte model has a heatsink with a raised heatsink so the thermal pad can be 1mm. Copper shims from aliexpress with paste between the shim and heatsink and thinner pads would give better thermal transfer.
I've got 3 meters roll of that for about 5€ from Ali ...that's cheap and works!
I would have added a test of the cheapest non-brand pads as well for comparison.
Great video Brian.
I've watched a few videos where they replace these pads with copper shims, and reduce temps around 20c or more on the 3080 & 3090 cards. Would be good to see you give that a try for comparison.
For my 5700xt the gelid extreme pads reduced about 10 to 15 C but my core temp rised a little . Expensive, worth it if youre watercooling
What temperature are you getting? I was getting about 70c while i was mining with the card which seemed fine to me. Kinda lazy to open the card but the thermal paste is due as well it's been 2 years since i bought it.
Even at 64 degrees, that's a long way off throttle temp for the G6X. I've always thought G6X ran really hot, but maybe it's just this cooler is really good or they used better pads? But if you're buying a used 3070 Ti on up... I might not consider it if they haven't done something for the memory since it's really easy to hit throttle temp when mining. I know there are a lot of better options, all the way up to copper shim plates that really help with cooling.
Good Advice. Thanks
I did a research as well back in the day about this like half a year ago... but I wasn't sitting on any High End card, I was using MSI Gaming X RX480 8GB for a quite some time, and I was pretty concerned about the thermal pads after changing my thermal paste, so I looked for something to replace them, just for the sake of getting peace of mind... and I did order the same pads that you did there but different thickness to match the originals.
After replacement my Memory Temp was slightly less... and core also dropped few degrees, so it didn't change much, but I think it was worth it anyway... I replaced my RX480 for RX6600 on my Main PC though, but I think I will do the same course of action with pads after some time anyway...
Used peosucts from Gelid. Good price/quality. I still got my Xtreme thermal paste and is easy to appky and works perfect.
I finally got around to changing my 3090fe out a couple days ago. I used the thermalrights (3 pads) 1.5mm and went from 104c to 80c. Can't believe how expensive this card is and had to do this to keep from nuking my system and prevent thermal throttling... Ridiculous... 😮💨
What case are you using 10:20? Looks Tech Yes Crackalackin
hopefully you could score some Aluminum Nitride Pads and test it out, I've read some articles online that it has the highest w/mk compared to the rest of the thermal pads being sold online
I'm really more curious how the stock ones compare to some really cheap 1 dollar pads, I've used plenty of those, and they always seem to do okay, but I have never properly tested them, I might give it a try on my 3080 since I have quite a few left
The sea hawk x may also be getting a new VBios with higher limit draw soon too
MSI are easily the most conscientious when it comes to the pads on their cards - try a pad swap on a Gigabyte or EVGA and the difference is much more marked.
Brian, try to have a look at the PTM7950 thermal interface material, because reddit recently started to go nuts about it
I thermal pad modded my 3080 FE with 2mm Gelid Extreme pads and saw some improvement. Two weeks later, I copper modded with 1.2mm plates ($10) and thermal paste on either side and the improvement was much more significant.
I don’t mine and cap my frame rate to 60fps while gaming. I just want my GPU to run cool and last a long time. My next rig will be in at least 5 years.
Imagine having a 3080 and capping your fps at 60
@@99mage99 well if it's 4k monitor, 60fps makes a lot of sense.
@@99mage99 My monitor is a content creation monitor and is at 60hz. More fps caused tearing issues so 60 fps is smooth enough for me.
@@t0nyxgq Imagine having a 3080 and only one monitor.
@@99mage99i have 2 monitors both at 60hz. But recently I got a third 2K monitor at 144hz and it’s very smooth. I use that for gaming now.
Where is the intel undervolting video workshop ?? And can you build a gaming rig what consumes less power than normal.
would love to see an updated xeon build and your opinion as if its worth it. cheers!
if you do it to your main gpu,we'd love to see another test to see if this was a one off situation or if it looks to be more or a trend
I got my brand new msi 3080 gaming trio 10gb. Came not new, not sealed. Definetly repackaged. I dare not return it though since i won it on a shuffle. After watching a tear down video i notice my cards missing thermal pads all over. Would love to repad and repaste it.
You also put bigger pads on there for more heat dissipation and transfer
5:08 I would not change the tape in this condition. And it seems to me that the tape that nVidia uses is much better, so two reasons I wouldn't go into this...
I used these same thermal pads to replace the Red Devil 5700XT stock thermal pads
Be curious to see the same test done but with the copper shims mod - use copper sheet the same dimensions as the thermal pads with the bare minimum of thermal paste applied between the memory and copper shim and then also between the copper shim and cooler.
I’ve seen this done in other vids but been dubious with the results - they seemed to good to be true.
Had similar problems on the VRMs of my ASUS Strix RX Vega 64. I changed the pads and the temps went down over 30 °C. Some GPUs have this problem.
2 other thing You could test, the Gel pads whit and without finger fat on it, and a copper mod.
Should've added kingpin paste on memory, then place the pad, then add more paste to finish the sandwich will give you better thermal transfer. Also, the gelid pads are too hard, so look for softer pads like Valor Odin.
I’ve been checking eBay here in Oz for used 3080 and 3080ti’s. At the moment there’s only a couple for sale. One local guy wants the same price for his mined card as a new one. I offered him aud$1000 and got no reply. I’m not sure we will see many mined cards on the market as these guys over invested when the prices were high and can’t come to part with the cards at a realistic second hand price.
lol that's their problem not yours. if he doesn't sell now he'll have a hell of a wake up call next week.
crypto bros are total clowns
@@ChrisZukowski88 yeah i could care less thats why i offered him 1000. pretty generous for a second hand mining card i thought? He will probably end up keeping it, I would espacially if i paid like 2500 for it lol
I've replaced mine on my 2070 with these, they're great it also looked like my specific card had some missing on a set of vram. I have also heard after a certain point you might have deminishing returns as the backplate cannot dissipate the heat quickly enough. Would recommed 10/10 mine dropped about 10C
Hey Bryan, love the videos keep them coming!!!! What thermal paste do you recommend/did you use on the 3080?
I have copper heatsinks on my bacck plate from when I was mining. I just placed them on with out opening it up.
I assume you changed thermal paste as well... Which one did you use and did it improve? Thanks
I had great success in that my VRM's didnt have pads so I added pads onto the VRM's and my GPU hotspot temperature decreased by about 10 degrees at full load.
I would have thought that we are loading the cooler with more heat so that would increase the GPU temp but actually no, for some reason,
the memory and gpu temps were great.
oh, I have a 6700XT Mech X2 OC and was having issues with driver timeouts before that.
I am thinking my next step may be to make the backplate a passive cooler by putting pads inbetween cos at the moment it is just for show.
I had this timeout with the msi rx 6600 mech 2X model too but the card had like 52 Celsuis max temp. What a weird model
The VRM is thermally well coupled to everything on the GPU with a lot of the coupling area having little opportunity to remove the heat, it's essentially forced out through the components. The coupling occurs via the massive ground and voltage plane pours in the inner layers of the PCB. Heatsink actually usually has a fair bit of thermal impedance between the different areas and it can absorb that extra heat no problem either.
Fujipoly are the tops.
Bryan you should check out the k5-pro it should give you better performance and there is no concern about the size/height
So, forgive me, I'm but a newbie at these things. But you compared a GPU with (possibly) dried out (and certainly not new) thermal paste with the same one with new thermal paste and pads?
Is this correct?
If that is the case, how do we know the effect of the pads?
Review the juicy stake tho!
On the VRAM; for 3080 & 3090 ... high wmk pads are great on nvidia's founders editions, & completely unnecessary on evga's ftw3 ultras & above.
The original pads on your 3080 look good anyhow and you are water cooled?. The FE 3080 had some nasty oily pads, and mem temps would easily hit 100C, especially in mining. That's why most people did the pad swap. For myself and my original FE 3080 swapping the pads made a huge diff to the VRAM (like over 20C). Incidentally GPU temps don't drop and could even go up a little as you are now dumping more heat into the heatsink.
You do need to make sure you get the right pad size as it can affect the GPU die mounting pressure. I (and others I know) have had good results with Thermalright pads.
Could you revisit this and add thermal pads on the back so that the backside of the memory contacts the backplate? And see how much difference it makes since gddr6 is flip chip and the hot side is at the back
if the back plate was not paded its going to be tricky as he will have to figure out the thinkness of the pads i know from seeing miners change pads yes changing the pads on the back plate does help too even on my 3060 that i redid it helped
Which tier thermal pads do you recommend from cheap to expensive? for value for money for performance 👍
Silver paste! It's been unbeatable for decades and there's a reason for that..
Fyi to people out there if you've bought gigabyte graphics cards this is definitely something that should be done as stock thermal pads leak oil and are known to cause issues or loose cooling performance.
@9:34 does i5-12400 have 20 Threads?
Should have tried or compared the cheaper Amazon thermal pads where you get 4 sizes for $10-$15
Different pads have different thermal conductivity. That's why some go for a premium.
@@Safetytrousers oh yea I get that just see what temps diff would be between regular and “premium” pads.
@@killerrf The original pads on this GPU would have been regular pads. They wouldn't have been designed to be actively bad.
They charge you more than founders edition to only give you crappy pads on them wow
I have 2 asus strix 1080ti's, is it worth changing the memory pads?
Brian can you pls do a b660 riptide vs MSI a pro
How to find out where are the Vrams ? if some thermal pads has been completely demolished ?
hey tech city which is better extreme vs ultimate pads wanna buy some and do you know which thickness for rtx2070 super fe and rx5700 xt also a fe model
I added thermal pads to my Gigabyte 3060Ti backplate. So the backplate actually does something now
I replaced my stock ps4 pads with fujipoly ultra extremes and while I haven't measured, it does seem to have helped transfer more heat to the mid-plate. I compensated by cutting some holes and adding some fans. Looks pretty hc but it I like it!
You have to be really careful with choosing the right thermal pads, my gpu died after a day with the new thermal pads.
That sounds more like a GPU problem if it hits high temperatures it should shut down it shouldn’t die. Unless the card it was already faulty. There’s a few good UA-cam videos that show you all these different brands the squishiness and thickness of each so you can determine the correct thickness. Some of the 2MM are squishier than others or Stiffer.
You had the wrong size?
@@swizzler8053 the thickness was right, but the original ones were more soft and more thermal conductive
If you mod something, you have to be prepared to buy it twice if it breaks, which begs the question, why didn't you just buy the better model, to begin with? People who mod are either rich, YOLOers, have had experience with other cheaper/expendable parts, or are UA-camrs with warehouses full of parts to work on.
@@eliashowe7115 dude many GPUs didn't have good paste or pads, even on top of the line GPUs.
Not to mention that weather + humidity condition during shipping / storage can also affect their quality.
Is this just for water cooled hybrid cards?
Holy crap I wish I had temps that good. My freakin vram is over 100C. :(
I have a zotac 1070 from 2017, I want to do paste and pads but I have 0 clue how think of thermal pads I need, how do I figure that out?
Can you test Gelid Ultimate vs Thermalgrizzly Minus Pad Extreme? Those TG pads are an upgraded version of their old Minus Pad 8, and they are ridiculoudly expensive. I think 3x the price of Gelids. I can't find anyone who's willing to test them. My GPU doesn't have thermal sensors on memories so I can't really do it myself.
Can you help me for the Gelid GP-Ultimate thermal pad thickeness for VRAM, Backplate and those weird row of blacks from my Gigabyte RX 5700XT 8GB OC? the stock pads were dry and not sticky anymore.
Think this is good for gaming laptops?
Add copper pads enjoy your gaming
For a top mounted radiator (for exhaust), do you recommend a push or pull configuration?
Currently it’s in a push configuration but it’s damn noisy and wish to quiet it down a bit
This is very dependent on the case you are using/ config of the other fans/radiator placement. Need more info for very accurate advice. In general tho positive pressure in the case is best
I don't think that the test was accurate ... cuz on the first one the thermal paste was dried and on the second one it was fresh. which could be the only reason for the temp change. correct me if I'm wrong
Could you double up 1mm pads if that is all that is available?
I've found the K5 pro does better job then any pad I can find. And if this hack scub is, I'm sure others do too.
Hi, What you think if i put thermopad on gpu prossoser not thermopast !!!