Dude, Copper is completely safe as long as it does not touch any other gpu components but Vram. I too have not just copper shims but the whole copper plate on my Vram, and no kapton tape applied at all. All cool, no issues at all.
i simply did a copper shim mod u can put copper shims on the vrams, with thermal glue tape. and depending on how thick ur copper shims are, u can either put a bit of thermal paste or a 0.5mm thermal pad on top, to connect it to ur main cooler.
A lot of cards had issue running in that position. They just run hotter. Vapor chamber design is not very bright. Initially all GPU could run like that which is same position as in servers. But now, like 15 year ago, they changed fin orientation vapor chamber design and heatsink which make them to run cooler in horizontal position . With all that they still push a trend with riser cables, like there is no tomorrow. Try horizontal position for your GPU and shove it in a well cooled case. Fractal Torrent Compact is a good air cooled case. Dust around power delivery is too much. Is strange that you look at the performance of you pads but, you ignore details like - dust is a fair thermal insulator and with the right humidity level in your room will throw some small shorts all over PCBs and you'll find yourself troubleshooting unexplained errors in any OS and waste your time greatly. Don't be afraid of copper shims, they can't conduct any electricity cause simply can't escape the mounting pressure you apply when you re assemble your card. And don't use individual squares for each module, is counterproductive. Some memory modules run hotter than others, proximity to GPU or power delivery, bad design plating or heatsinks, faults in vapor chambers, bad plating and inconsistency. Basically you want to disperse the heat as fast as you can. Simply you have a bigger surface of copper or ceramic responsible to take the heat from VRAM and send it to the heatsink. In fact, with little squares is higher risk to escape the mounting pressure than with let's say, a shim that covers 4 modules. I've done my self several mods with copper shims and plates, even changed the stupid back plates of plastic or aluminum painted/ stickers and other crap with whole bare copper backplates. 8-15 C lower on VRAM. Never shorted a single card. 100 C I consider critical on VRAM specially where you have Hynix and Micron chips. Samsung chips tend to be more heat resistant but, is not enough info to staple that. I really don't care whatever manufacturer is saying that 110 C is safe on VRAM they never say for how long, 1 month, 4 months, 24/7 use or less?? IS plain BS!!!! the number of cards with artifacts I've seen, is sickening, caused by VRAM is huge specially starting with 2000 series onwards. Client is calling when is nothing can be done anymore because VRAM is gone 1 chip or 6 chips , who cares, card is dead or artifacts. The problem is why client did not called? - he believed the manufacturer >>>> 100 is safe or 110C is normal. Now the manufacturer sells another card. Work as intended. Hop[e it helps
Once you went too far with sanding your ceramic shims you could have swapped the thermal paste for thermal putty, putty is actually meant to fill gaps so you can use an excess and just use the pressure of screwing the cooler back on to squeeze out all the putty that isn't needed.
I would use a single sheet/shim and use U6 Pro putty on both sides, this way there is room for expansion otherwise memory chips can crack on the solder pads
instead of changing the thickness of the pads, what if you modified the heatsink to squeeze snuggly on to the pads with some springs under small tension? Aluminum can be a lot easier to work with.
Or just water cool your graphics card with a full-coverage block and correct pads. Save yourself the risk of cracking solder joints under chips, and keep stock cooler minty-fresh, for if you want to resell your card later. Plus the buyer starts off with a newly repasted GPU.
why not instead get some good thermal pads like the minuspad 8 and call it a day? yes it wont be as cool as copper or ceramic but defenately better than stock
People getting sick, buying a GPU card and then becoming prick about number, temps, tweaks or whatever, wasting the time on these things instead of the purpose the GPU was designed for….PLAY GAMES. That’s make no sense
It's not about numbers, it's about what they represent. If, in this case, mem temps hit 110 degrees, cards starts to thermal throttle, drop performance, fans start spinning like crazy. So it's not about number, it's about getting maximum from the thing you bought making the card last as long as possible while having a very quiet and silent in some scenarios card. Fan is literally does not spin. The card will last much longer and experience with the computer will be much more pleasant. I just don't wanna hear a sound of a vacuum cleaner in front of me.
nvidia failed at designing these cards. this is a fix for the cards' shortcomings and is not a waste of time. The life of the card will be a lot longer if it isn't destroying itself.
Dude, Copper is completely safe as long as it does not touch any other gpu components but Vram.
I too have not just copper shims but the whole copper plate on my Vram, and no kapton tape applied at all. All cool, no issues at all.
Not to mention that all decent cpu/gpu coolers have copper base plates. And yet regular guys on TH-cam think they know better.
@@darkoz1692 they do know better, because it helps to lower vram temperatures drastically.
i simply did a copper shim mod
u can put copper shims on the vrams, with thermal glue tape.
and depending on how thick ur copper shims are, u can either put a bit of thermal paste or a 0.5mm thermal pad on top, to connect it to ur main cooler.
A lot of cards had issue running in that position. They just run hotter. Vapor chamber design is not very bright. Initially all GPU could run like that which is same position as in servers. But now, like 15 year ago, they changed fin orientation vapor chamber design and heatsink which make them to run cooler in horizontal position . With all that they still push a trend with riser cables, like there is no tomorrow.
Try horizontal position for your GPU and shove it in a well cooled case. Fractal Torrent Compact is a good air cooled case.
Dust around power delivery is too much. Is strange that you look at the performance of you pads but, you ignore details like - dust is a fair thermal insulator and with the right humidity level in your room will throw some small shorts all over PCBs and you'll find yourself troubleshooting unexplained errors in any OS and waste your time greatly.
Don't be afraid of copper shims, they can't conduct any electricity cause simply can't escape the mounting pressure you apply when you re assemble your card. And don't use individual squares for each module, is counterproductive. Some memory modules run hotter than others, proximity to GPU or power delivery, bad design plating or heatsinks, faults in vapor chambers, bad plating and inconsistency. Basically you want to disperse the heat as fast as you can. Simply you have a bigger surface of copper or ceramic responsible to take the heat from VRAM and send it to the heatsink. In fact, with little squares is higher risk to escape the mounting pressure than with let's say, a shim that covers 4 modules.
I've done my self several mods with copper shims and plates, even changed the stupid back plates of plastic or aluminum painted/ stickers and other crap with whole bare copper backplates.
8-15 C lower on VRAM.
Never shorted a single card.
100 C I consider critical on VRAM specially where you have Hynix and Micron chips. Samsung chips tend to be more heat resistant but, is not enough info to staple that.
I really don't care whatever manufacturer is saying that 110 C is safe on VRAM they never say for how long, 1 month, 4 months, 24/7 use or less?? IS plain BS!!!! the number of cards with artifacts I've seen, is sickening, caused by VRAM is huge specially starting with 2000 series onwards. Client is calling when is nothing can be done anymore because VRAM is gone 1 chip or 6 chips , who cares, card is dead or artifacts. The problem is why client did not called? - he believed the manufacturer >>>> 100 is safe or 110C is normal. Now the manufacturer sells another card. Work as intended.
Hop[e it helps
Once you went too far with sanding your ceramic shims you could have swapped the thermal paste for thermal putty, putty is actually meant to fill gaps so you can use an excess and just use the pressure of screwing the cooler back on to squeeze out all the putty that isn't needed.
th-cam.com/video/WswfRRGoJKE/w-d-xo.html
I would use a single sheet/shim and use U6 Pro putty on both sides, this way there is room for expansion otherwise memory chips can crack on the solder pads
5:45 why did you not use new good flexible thermal pads instead?
you would also need to consider the extension of the parts under heating condition
Which makes the idea even more stupid
that jawline is sharper than ma kitchen knife damn
Not good for boxing
instead of changing the thickness of the pads, what if you modified the heatsink to squeeze snuggly on to the pads with some springs under small tension? Aluminum can be a lot easier to work with.
i have more than 9 3080's and 3090 all copper modded day one, they aren't isolated in any way, i use thermal putty + copper shims, lovely
Defeats the purpose if you gotta stack pads.
Or just water cool your graphics card with a full-coverage block and correct pads. Save yourself the risk of cracking solder joints under chips, and keep stock cooler minty-fresh, for if you want to resell your card later. Plus the buyer starts off with a newly repasted GPU.
tl:dr watch his other video cos this didnt work
why not instead get some good thermal pads like the minuspad 8 and call it a day? yes it wont be as cool as copper or ceramic but defenately better than stock
th-cam.com/video/WswfRRGoJKE/w-d-xo.html
thx when i get a gpu that even has a the coler contact the memory il use that(i have a 1660tirn)
make a ceramic mold , under high pressure
i raise you a shot glass of vodka and scrap the ceramic pads , i am using 99% pure Silver pads !!!! works great 78c on the memory !!!
dude why u reading off the wall? say your own opinion bruh and coper modding isn't risky why u even care? it's his card, not yours so
You sound 100% like Latka XD. You know? That Andy Kaufman character
Why did you have to say this?
please work on better presentation. Stop excessive eye movements. and i had to speed up 1.5x to focus on the subject
People getting sick, buying a GPU card and then becoming prick about number, temps, tweaks or whatever, wasting the time on these things instead of the purpose the GPU was designed for….PLAY GAMES. That’s make no sense
It's not about numbers, it's about what they represent. If, in this case, mem temps hit 110 degrees, cards starts to thermal throttle, drop performance, fans start spinning like crazy. So it's not about number, it's about getting maximum from the thing you bought making the card last as long as possible while having a very quiet and silent in some scenarios card. Fan is literally does not spin. The card will last much longer and experience with the computer will be much more pleasant. I just don't wanna hear a sound of a vacuum cleaner in front of me.
you are a low iq citizen, congrats
nvidia failed at designing these cards. this is a fix for the cards' shortcomings and is not a waste of time. The life of the card will be a lot longer if it isn't destroying itself.
@@goofyb6951 make sense in this case, ya right!
In other hand, there are people crazy about numbers too (when everything is running “normal”).
Keep doing this, so your GPU's break and they are sent to people like me to fix them :D