Artifacts after cleaning gpu

[soundopossum]

• So, i have a gtx 1650 by gigabyte and i wanted more performance for no cost so looked over to over clocking and i was genuinely surprised i was able to get a max core clock of 1980 mhz (stock is 1410mhz) and a plus 740 on my memory with no artifacts but temps were high and i thought well i should probably replace the thermal paste but when i opened uo my gpu there was thermal paste everywhere execpt the die and it was everywhere surrounding the die so i soaked only the die in isopropyl alchohol then skip forward 20 min i first used a q tip then to get the capacitors surrounding the die i ....... Used a toothpick i didnt poke at them i just used it to get around them then i used a microfiber cloth ( some of u might say it creates static but these gpu have grounding ports now after reassembling the gpu these things happened.
• Any memory overclock over plus 400 causes artifacts and it crashes
• temps are worse by three degress  

Core clock is ok and im using the same curve but for some reason my memory is fcked gpu z still does recognize the sad 4 gigs i have of vram 

[DoctorNick]

2 minutes ago, soundopossum said:

• So, i have a gtx 1650 by gigabyte and i wanted more performance for no cost so looked over to over clocking and i was genuinely surprised i was able to get a max core clock of 1980 mhz (stock is 1410mhz) and a plus 740 on my memory with no artifacts but temps were high and i thought well i should probably replace the thermal paste but when i opened uo my gpu there was thermal paste everywhere execpt the die and it was everywhere surrounding the die so i soaked only the die in isopropyl alchohol then skip forward 20 min i first used a q tip then to get the capacitors surrounding the die i ....... Used a toothpick i didnt poke at them i just used it to get around them then i used a microfiber cloth ( some of u might say it creates static but these gpu have grounding ports now after reassembling the gpu these things happened.
• Any memory overclock over plus 400 causes artifacts and it crashes
• temps are worse by three degress  

Core clock is ok and im using the same curve but for some reason my memory is fcked gpu z still does recognize the sad 4 gigs i have of vram 

Try again, also no toothpick poking this time. Just wipe with paper towel and clean the remainder with some IPA and paper towel. Make sure thermal pads is sitting correctly and isn't broken. Be careful when tightening the screws. Tighten the screws around the GPU die in a x-pattern. 

[soundopossum]

9 minutes ago, DoctorNick said:

Try again, also no toothpick poking this time. Just wipe with paper towel and clean the remainder with some IPA and paper towel. Make sure thermal pads is sitting correctly and isn't broken. Be careful when tightening the screws. Tighten the screws around the GPU die in a x-pattern. 

Ok but i doubt that will help as i did everything correct (expect toothpick poking) including tighning in a x pattern i had thermal ppads on but they made no difference to temps so i choosed to remove them to make it easier if you need an image i would be glad to send

[DoctorNick]

4 minutes ago, soundopossum said:

Ok but i doubt that will help as i did everything correct (expect toothpick poking) including tighning in a x pattern i had thermal ppads on but they made no difference to temps so i choosed to remove them to make it easier if you need an image i would be glad to send

Do you know the temp of your VRM and VRAM before and after. I would guess thats the reason the card isn't stable anymore.

[soundopossum]

2 minutes ago, DoctorNick said:

Do you know the temp of your VRM and VRAM before and after. I would guess thats the reason the card isn't stable anymore.

Sadly the gtx 1650 dosent have sensors for the vrm or the vram but it ran just fine with the overclock for 2 years

[DoctorNick]

3 minutes ago, soundopossum said:

Sadly the gtx 1650 dosent have sensors for the vrm or the vram but it ran just fine with the overclock for 2 years

So the thermal pads have been removed for two years? Then idk what the issue is

[soundopossum]

So, i have a gtx 1650 by gigabyte and i wanted more performance for no cost so looked over to over clocking and i was genuinely surprised i was able to get a max core clock of 1980 mhz (stock is 1410mhz) and a plus 740 on my memory with no artifacts but temps were high and i thought well i should probably replace the thermal paste but when i opened uo my gpu there was thermal paste everywhere execpt the die and it was everywhere surrounding the die so i soaked only the die in isopropyl alchohol then skip forward 20 min i first used a q tip then to get the capacitors surrounding the die i ....... Used a toothpick i didnt poke at them i just used it to get around them then i used a microfiber cloth ( some of u might say it creates static but these gpu have grounding ports now after reassembling the gpu these things happened.

Any memory overclock over plus 400 causes artifacts and it crashes

temps are worse by three degress  

Core clock is ok and im using the same curve but for some reason my memory is fcked gpu z still does recognize the sad 4 gigs i have of vram 

[leclod]

Removing thermal pads is a criminal offense.

[Poinkachu]

1 hour ago, soundopossum said:

• when i opened uo my gpu there was thermal paste everywhere execpt the die and it was everywhere surrounding the die
• temps are worse by three degress  

 

There's a chance they used Thermal Putty instead of thermal pads for anything other than the die. It's a substance more viscous than paste (almost like a playdoh) but sure can look like a somewhat dry paste.

 

You can replace it with pads, but the thickness have to be right or you'll either create gap between the die+paste and the heatsink, or everything that you put pads on doesn't touch the heatsink. Be it slightly or at all.

 

Recent GPU example that uses Putty that I can remember right now would be EVGA RTX3080ti FTW3

 

And OP, there is no need to create 2nd topic asking about the same thing.

[soundopossum]

yeah but i fixed the problem i simply falshed my drivers and now it works?

[Wardus]

Did you replace all of the thermal pads? Did you clean the heatsink as well as the board? It is possible that you knocked something off of a power management system, and when the card ramps up and starts channeling power a certain way, it gets unhappy. I'm going to guess that a mistake was made somewhere. I was cleaning the back of a 2070 Ti a while back and knocked one of the resistors for a PCIe trace off and bricked the card :(. Just stick with normal settings and you'll be fine. Probably all you can do.

[LogicalDrm]

*** Threads merged ***

 

Please do not post multiple times about same thing.

[soundopossum]

3 minutes ago, LogicalDrm said:

*** Threads merged ***

 

Please do not post multiple times about same thing.

Sorry, i wont do it again