• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

How to stop overheating integrated GPU?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Wolfdawg

Registered
Joined
Apr 1, 2012
Location
Cossville, Tennessee
When I play a very graphic intensive game (Crysis) my PC will instantly shutdown after a few minutes, while playing a not so graphic intensive game (MW2) my PC will last forever or 1 hour+. The only part seeming to cause the shutdown is the integrated GPU on my Motherboard, its the only hot part on my Motherboard, depending on the game my GTX560 will only get warm but the Chipset very hot.

I don't have any temperatures for the GPU but it is hot to the touch when playing a game. I've singled it out as the Chipset because I was watching 1080p anime with MadVR as the render-er (GPU intensive) and my PC shut down, my GTX560 and CPU heatsink were slightly above room temp but the Chipset was burning hot.

This all started after 2 MOSFET's blew on my motherboard and a underclock to 2.4ghz, I'm working on replacing them or a RMA.

How do I stop this Integrated GPU from overheating? Disabling it in the BIOS did not work and I get 0% usage when I use only it. (take out my GTX560)
 
If you have a 560 then the integrated gpu should not be being used at all during gaming.

I think your problem might be related to the motherboard. You should get on RMA'ing it or buying a new motherboard and your problems will probably go away.

What motherboard/case/psu/other parts are you using in this system?
 
The Chipset not only has to do with the integrated GPU but the PCI-e bus as well and so it is involved when your pushing your GTX560 which uses the PCI-e bus. That's why it's getting so hot.

You give us no information about your motherboard make and model. If it is an inexpensive motherboard, stuff on it will run hot because 1. the heat sinks on the chipset are small, 2. the power phase is not heavy duty (probably 4+1 instead of 8+2, for instance) and the mosfets and other power control components aren't sinked at all. Plus, if it's a mATX board instead of a full ATX the smaller size inhibits heat dissipation and crams hot components closer together.

You can always look into putting bigger sinks on the chipset or cooling them with fans. If you blew mosfets, that's a sure sign the board was getting way to hot, way overloaded.

What is the make and model number of the board? What is the board's TDP rating? A Phenom II X4 will pull a lot of watts when overclocked to 4.0 ghz.

What is the make and model of your case. How's the ventilation? How many intake and exhaust fans do you have in the case?

Have you actually done any temp monitoring while gaming? You should have HWMonitor open on the desktop while you game. It will record max CPU, core and motherboard termps. Feeling the CPU heatsink while it is loaded and saying "it was not much hotter than room temp" is not a good way to evaluate temps. If the heatsink is big and efficient it will dissipate heat so fast it won't build up much but the socket area could still be quite warm.
 
What is the make and model number of the board? What is the board's TDP rating? A Phenom II X4 will pull a lot of watts when overclocked to 4.0 ghz.

What is the make and model of your case
Additionally, what is the make and model of your power supply?
 
The Chipset not only has to do with the integrated GPU but the PCI-e bus as well and so it is involved when your pushing your GTX560 which uses the PCI-e bus. That's why it's getting so hot.

You give us no information about your motherboard make and model. If it is an inexpensive motherboard, stuff on it will run hot because 1. the heat sinks on the chipset are small, 2. the power phase is not heavy duty (probably 4+1 instead of 8+2, for instance) and the mosfets and other power control components aren't sinked at all. Plus, if it's a mATX board instead of a full ATX the smaller size inhibits heat dissipation and crams hot components closer together.

You can always look into putting bigger sinks on the chipset or cooling them with fans. If you blew mosfets, that's a sure sign the board was getting way to hot, way overloaded.

What is the make and model number of the board? What is the board's TDP rating? A Phenom II X4 will pull a lot of watts when overclocked to 4.0 ghz.

What is the make and model of your case. How's the ventilation? How many intake and exhaust fans do you have in the case?

Have you actually done any temp monitoring while gaming? You should have HWMonitor open on the desktop while you game. It will record max CPU, core and motherboard termps. Feeling the CPU heatsink while it is loaded and saying "it was not much hotter than room temp" is not a good way to evaluate temps. If the heatsink is big and efficient it will dissipate heat so fast it won't build up much but the socket area could still be quite warm.

Sorry, didn't think this needed that much detail.

Its a Micro ATX board, MSI: NF725GM-P43 (can't find the TDP rating)

Like I said this all started after the MOSFET's blew and not from heat. (Pretty sure the MOSFET's didn't die from heat, but did blow from overload)

as far as ventilation I've played Skyrim On very high settings (almost maxed) at 102f, 3.40ghz. When I had my PC by the A/C.


Possibly its putting more strain on the Chipset the lower my CPU clock? I've never underclocked it so I wouldn't know. I can't go any higher than 2.40ghz because of the blown chips, while I was watching the 1080p anime my CPU usage was at 30% so it shouldn't have been the reason it turned off but the Chipset overheating fault.
 
Man if you're trying to figure out what's happening with a board that has two blown mosfets, what more do you need to look for? You put a PI X4 on a low TDP board. Probably okay to do if the CPU was in the support list for that board but not okay when you overclock it. If you're going to overclock a CPU make sure the CPU TDP is less than the TDP rating of the board to start with. Overclocking drives the TDP of the CPU way up.
 
Man if you're trying to figure out what's happening with a board that has two blown mosfets, what more do you need to look for? You put a PI X4 on a low TDP board. Probably okay to do if the CPU was in the support list for that board but not okay when you overclock it. If you're going to overclock a CPU make sure the CPU TDP is less than the TDP rating of the board to start with. Overclocking drives the TDP of the CPU way up.

I didn't mention that the CPU was not overclocked or at 4.0ghz, other than once to test. According to a prim95 test the stability of my PC was fine, I then clocked it back to factory because this board sucks and I had to restart the mobo every boot because the multiplier would reset.

So are you sure that's what is wrong with my PC, the MOSFETs? (causing the chipset to overheat, cuz seriously it gets ****ing hot while playing crysis)
 
I think it's fair to say that buying a new motherboard will most likely solve your problems. Even without the blown MOSFET's that board is cheap, but it's definitely the most obvious cause of your problems.

Assuming you're Phenom x4 CPU is a 955, 965 or 960, something like this should do wonders in terms of stability and overclockability.
 
MSI says the NF725GM-P43 supports 125W CPUs, and generally doesn't CMOS digital chip dissipation go up proportionally with clock speed, meaning a CPU rated for 95W @ 3.0GHz will dissipate 127W @ 4.0GHz?

Are MSI motherboards more prone to failure from overclocking than other brands are, maybe because they're built more "economically"? I noticed that my MS-7309 v. 2.3 socket AM3 mobo has 3 spots where MOSFETs were left out, and some spots for tiny resistors just have solder blobbed over them. A few fuses and overvoltage protection diodes are also empty.
 
If you don't increase the voltage, a clock speed increase will linearly increase heat. Increasing voltage is exponential.
 
So since I bought Black ops a few days ago, I hadn't the use to do this until now. When I started the game and maxed everything out at 2.0ghz, I got some crappy frames and found out Black ops is very CPU intensive and was only using 30% of my GPU.

I lowered my CPU voltage to 1.2 and clocked it at 2.6ghz, since then my PC hasn't shutdown from playing a game or bluescreen (although it did bluescreen once when the voltage was too low)

I've clocked it to 2.7ghz now and will continue until I find the max stable. Black ops is working fine and other games are working better.
 
Back