Motherboard abnormally hot under load

[notbike]

So after switching out my dead GPU for a much better one, my temps have drastically changed, especially my motherboard.

When I had the AMD Radeon 7950 my idle temps looked like this:
CPU 35-40ºc
Motherboard 30-35ºc
GPU ~30ºc

Under moderate to high load the CPU would rise to 60, the motherboard would hit maximum 40 from I guess the ambient temperature rise in the case, and the GPU could hit up to 70 but never really over 70.

Since installing the GTX 980Ti my idle temps have all increased by around 5 except the GPU temp which is currently at 28ºc (Even that doesn't add up).

I don't mind the idle temps having increased a little, but what I find alarming is that my motherboard has shot up by 30ºc when under moderate/high load since the GPU switch.

When on maximum GTA5 graphics settings at 60fps (Frame scaling off) my GPU will hit 80 after a while. CPU stays around low 60s, but the motherboard has hit 74 at one point which was alarming. My friend's motherboard during the same gaming session on maximum graphics with a GTX 980 stayed at a solid 32ºc, while other friends reported not going above 40.

When I checked temps in OpenHardwareMonitor it gave me 3 readings from 3 motherboard sensors. This was immediately after closing GTA5 so temps were falling again. Two sensors read 33 and 34, but one of them was reading low 60s.

No idea where that sensor was placed, but even if it was near the GPU, the GPU was already falling below 40s at that point, and the CPU was falling to low 50s. It can't have been another component causing it to read almost twice the temp of the other motherboard sensors. I'm thinking either the sensor is dodgy or a part of my motherboard is getting very hot.

tl;dr Why is my motherboard heating up so much when install new gfx card?

Also as a side note, the PC was spotless at the time of taking these temps so it's not dust and I have a decent push/pull airflow with plenty of fans.

[Aff]

Surely you have answered your own question?

You have said that you have bought a more powerful GFX card that is running at lower temps, meaning that it is doing a better job of getting rid of the extra heat than you last one.

You've also said that its just one sensor that's getting hot.

So it seems quite likely that your card is dumping a lot of hot exhaust onto the area where one of the thermistors is, making it look like the board is hot. I would also imagine there is a chunky ground plane or something near by retaining the heat for a lot longer than the GFX card, which is made to get back to ambient ASAP.
____________________
Current Bikes:Honda 929RR Fireblade, Honda CD200 Benly (Project), Stomp Z2 140
Electric Bike Project

[notbike]

[CaNsA]

What motherboard?

[notbike]

[CaNsA]

Righto,

Drop a stick of ram, test each one individually.

If that turns up nothing then take the whole rig to a mates house and test out each bit.
GFX
CPU
RAM
PSU

If they are all good then start testing your motherboard with the other bits.
Add in his RAM, CPU ( if possible), GFX and PSU.



I reckon the PSU is buggered, or possible the ram.

[notbike]

[Accuhaler]

[CaNsA]

If the old GFX popped, who knows what has happened to the other parts...

[Accuhaler]

[CaNsA]

[temeluchus]

[temeluchus]

[notbike]

[temeluchus]

the problem could be that the new card could be venting heat right onto the sensor. Or any number of things.

12v is only nominally 12v it will fluctuate.
____________________
Some shite cruiser. Now with guns and FREEDOM!

[orac]

I haven't heard anything great about alienware (owned by dell, and their kit isn't exactly great) stuff. CPU was running a bit hot to start with.

I would suspect that the is a regulator issues on the motherboard, most likely a heat sink not seated correctly. although the GPU will get most of its power from its ATX connector some power will come from the PCIe port and thusly the mother board has to supply it.

what CPU are you using. I personally cant see the ram making a huge difference as pretty much everything in the past 10 years or so are using IMC so nothing other than the RAM and CPU are involved there.

I would have a good look round and make sure you didn't nock a heat sink while installing the new card causing it to become unseated.

the rest of the machine specs may well be helpful and if you can with the help of a friend if needs be get some temperature readings of the heat sinks so we can work out if it actually getting that hot or just a faulty sensor. @ 60 degree even an incorrectly seated heat sink will get pretty hot.
____________________
Current rides - 2016 Triumph Street Triple Rx, 1994 Suzuki Bandit 400 VM, TGB 204 Classic 125cc
"with nothing left to lose, there is everything to gain. It's not the size of the dog in the fight, it's the size of the fight in the dog"

[prawny1]

did you disturb the chipset heatsinks? they use different themal interface material than cpu's etc, generaly they are a fit once phase change material that sets when cooled so if you disturb the bond the temps will never be the same.

I sometimes remove the chipset heatsinks flat the mating surface to improve contact area and apply decent thermal adhesive to re secure them.

One I found to work ok is the halnziye heatsink plaster it actually dried flexible a bit like silicone and improved temps a fair bit, it works great on laptop gpus for bonding shims in place.

Another thing to consider is that the gpu being more powerful can allow for greater bandwidth/ through put on the other components.

Are you using the dual graphics feature on your motherboard? have you cleaned and updated your drivers to make sure no software is conflicting?
____________________
hellkat wrote:

Some people are only alive cos its against the law to kill them.