Question my gpu temp rises in sec

[goolakh]

hi my gpu temp rises from 40to 85c in max 10 sec when using furmark i check thermal paste and its fresh i check heat sink it was cold but chip Some warm . ur think its problem from heat sink(or heat pipe) or chip have problem?maybe temp sensor my gpu is r9 280x sapphir 3gb

 

[Fix_that_Glitch]

Are the cooling fans working? Do you have them set to a higher fan curve to cool them? You can use msi Afterburner to control the fan speed on it.

 

[Jmi20]

Based on my experience with an r9 380, heavy stress tests tend to make these older cards heat up a lot. The only way i was able to tame the temps was by undervolting.

But of course theres no harm checking the basics. You say that the heat sink is cold but temps are high? assuming you replaced the thermal paste, did you tighten the screws properly? If that's the case it might be the software not reporting temps properly, what software do you use to check temps?

 

[dorsai]

85c for a GPU being hammered with a load it will never see in normal use is not bad at all, especially for an older card like a 280x which was hot to begin with. I'm not a fan of hammering on older GPUs with Furmark myself...unless you want to kill the card. A better way to test a cards thermals is to loop an actual game benchmark and monitor temps IMHO but I'm sure others feel differently.

 

[goolakh]

Are the cooling fans working? Do you have them set to a higher fan curve to cool them? You can use msi Afterburner to control the fan speed on it.

one fan some slower than another so i place 2 more fan behind and front and idont think this is about fan problem i will test after that tell u result ty

 

[goolakh]

Based on my experience with an r9 380, heavy stress tests tend to make these older cards heat up a lot. The only way i was able to tame the temps was by undervolting.

But of course theres no harm checking the basics. You say that the heat sink is cold but temps are high? assuming you replaced the thermal paste, did you tighten the screws properly? If that's the case it might be the software not reporting temps properly, what software do you use to check temps?

It has two screws
I had seen the sides of the chip on the capacitor or the exact resistance, I do not know, the dough was spilled, can it cause a problem?

 

[goolakh]

85c for a GPU being hammered with a load it will never see in normal use is not bad at all, especially for an older card like a 280x which was hot to begin with. I'm not a fan of hammering on older GPUs with Furmark myself...unless you want to kill the card. A better way to test a cards thermals is to loop an actual game benchmark and monitor temps IMHO but I'm sure others feel differently.

first i test in dota 2 , after 10-15 sec in dota menu temp rises 80c-85c its not normall
i use gpu-z

 

[Fix_that_Glitch]

use cpuid hwmonitor and see what kind of temps it tells you. If it also results in high temps then you know it is running hot.

 

[Jmi20]

It has two screws

I think this may be the problem. The gpu cooler must always have the four screws secured, else the gpu die will not make good contact with the heat sink.

 

[goolakh]

iwill test after fix screws and tell u result

 

[goolakh]

I think this may be the problem. The gpu cooler must always have the four screws secured, else the gpu die will not make good contact with the heat sink.

i hope this is the problem. i will buy screws and after fix tell u result

 

[goolakh]

image
look at my gpu chip they use thermal paste dirty(Sides of chip ) its bad for temp?
they use black glue on heat sink for black thing on gpu board is it default?

 

[David0ne86]

2xx and 3xx serie cards are known for running in the 80s. It's completely normal. I have a 290x in my backup rip and it reached 85 degrees unless undervolted. You should try to undervolt as it's very easy using the adrenalin software.

EDIT that's WAY too much paste. It's pretty much smeared all over. Too much is just as bad as too little.

 

[Jmi20]

I don’t mean to start an argument but i tend to disagree when someone says “too much thermal paste”. Any excess would just be squeezed out. The worst that would happen is you make a mess and waste thermal paste

the real danger with a bare die such as a gpu is too little. If any part of the die is not covered with thermal paste or is not making good contact there may be hot spots which will not get properly cooled.

OP, If what you say regarding the screws is accurate, then the screws are the solution to your problem. The gpu will not make good contact with the cooler with only two out of the four screws.

beware that the screws of the cooler arent regular screws, they are spring loaded.

 

[Fix_that_Glitch]

I don’t mean to start an argument but i tend to disagree when someone says “too much thermal paste”. Any excess would just be squeezed out. The worst that would happen is you make a mess and waste thermal paste

the real danger with a bare die such as a gpu is too little. If any part of the die is not covered with thermal paste or is not making good contact there may be hot spots which will not get properly cooled.

OP, If what you say regarding the screws is accurate, then the screws are the solution to your problem. The gpu will not make good contact with the cooler with only two out of the four screws.

beware that the screws of the cooler arent regular screws, they are spring loaded.

I agree the only danger of too much thermal paste is if it gets on other parts of the card/motherboard. There are plenty of tests online with pc sites test show pea size, x across,too much, doesn't really matter for cooling effiency as long as there is thermal paste covering the heat generating zones. So I don't think anyone has asked what is the temps during regular use?

 

[Jmi20]

To give a better explanation for op and to make sure we are on the same page:

circled in red are the screws i am talking about. From what i understand from your post only two out of those four screws are installed. You must make sure those screws are complete. That is the cause of your overheating.

i assume the “black thing” op is talking about are the ram chips, ive drawn yellow marks on those. They use thermal pads, not glue. I recommend replacing them with any thermal pad with the same thickness.

forgive the sloppy editing of the photo, i did this on my phone.

 

[Fix_that_Glitch]

To give a better explanation for op and to make sure we are on the same page:

circled in red are the screws i am talking about. From what i understand from your post only two out of those four screws are installed. You must make sure those screws are complete. That is the cause of your overheating.

i assume the “black thing” op is talking about are the ram chips, ive drawn yellow marks on those. They use thermal pads, not glue. I recommend replacing them with any thermal pad with the same thickness.

forgive the sloppy editing of the photo, i did this on my phone.

I was going to mention thermal pads, I am sure they are very tired by now.

 

[Fix_that_Glitch]

Also use cpuid Hwmonitor. It will tell you if it is your gpu or memory that is getting too hot. Then you will know what needs addressing if not both. HWMONITOR | Softwares | CPUID

 

[dorsai]

I don’t mean to start an argument but i tend to disagree when someone says “too much thermal paste”. Any excess would just be squeezed out. The worst that would happen is you make a mess and waste thermal paste

That depends entirely on the thermal paste in question...some of it is conductive and some GPU dies have electrical components mounted in close proximity to the die which is a bad combination for GPU life.

 

[goolakh]

2xx and 3xx serie cards are known for running in the 80s. It's completely normal. I have a 290x in my backup rip and it reached 85 degrees unless undervolted. You should try to undervolt as it's very easy using the adrenalin software.

EDIT that's WAY too much paste. It's pretty much smeared all over. Too much is just as bad as too little.

temp +40c rises after 10sec in game or furmark! not normall

 

[goolakh]

I agree the only danger of too much thermal paste is if it gets on other parts of the card/motherboard. There are plenty of tests online with pc sites test show pea size, x across,too much, doesn't really matter for cooling effiency as long as there is thermal paste covering the heat generating zones. So I don't think anyone has asked what is the temps during regular use?

before installing drivers 55-60C after that 45C without using just run windows programs

 

[goolakh]

To give a better explanation for op and to make sure we are on the same page:

circled in red are the screws i am talking about. From what i understand from your post only two out of those four screws are installed. You must make sure those screws are complete. That is the cause of your overheating.

i assume the “black thing” op is talking about are the ram chips, ive drawn yellow marks on those. They use thermal pads, not glue. I recommend replacing them with any thermal pad with the same thickness.

forgive the sloppy editing of the photo, i did this on my phone.

yeah its have 2 screws behind red circles
about yellow : idont know maybe its thermal pads, When buying screws, I tell him check
no problem its so good <3

 

[goolakh]

I was going to mention thermal pads, I am sure they are very tired by now.

In case of failure, is it possible for the graphics temperature to increase significantly in a few seconds?

 

[Fix_that_Glitch]

In case of failure, is it possible for the graphics temperature to increase significantly in a few seconds?

Yes so use cpuid hwmonitor. It will tell you if the gpu chip is over heating, or if it is your GPU memory, or both. Then tell us what it says.

 

[goolakh]

some of thermal paste sides of chip, i should will cleaning that?
image