FX-9590 Temperatures too High with Factory FX-8150 Water Cooler

[Ravenas]

I have been using an FX-8350 with a FX-8150 water cooler for about 2 years. Recently I was able to get a FX-9590 for free. The FX-8150 factory AMD water cooler by Astek that I have isn't able to cool my current FX-9590. My computer freezes every time I play a game for 15 minutes, approximately. I ran OCCT stress test several times on my computer, and every time my processor climbs to 60 degrees Celsius and then it freezes (no blue screen and not crash text). Is this water cooler not sufficient for the FX-9590? AMD used to sell the exact same water cooler I have with the FX-9590 so I don't understand... Am I going to have to invest in a different water cooler or is this a problem with the processor?

 

[OneMoar]

should't be locking up at a only 60c temps are not your problem here
you have a problem elseware
I would guess power supply or board try clearing the cmos and ensure your bios supports the 9590
bios version 1403 or higher is required for the 9590

 

[Ravenas]

should't be locking up at a only 60c temps are not your problem here
you have a problem elseware
I would guess power supply or board try clearing the cmos and ensure your bios supports the 9590
bios version 1403 or higher is required for the 9590

I have version 2002. My power supply should be plenty... Gold certified 850W Corsair. It happens EVERY time at 60 degrees C or higher. I'm not overclocking at all. The processor is at factory settings. Same results when I turn turbo core off.

 

[Silas Woodruff]

I'm guessing 60 degrees Celsius would be lower than the maximum temperature the CPU can operate so it might be a CPU defect, if otherwise it works fine till 60 degrees investing in a hefty cooler might fix the problem.

Edit: http://forums.anandtech.com/showthread.php?t=2343240 in this thread the op claims his CPU runs at 40 degrees with the stock cooler but I haven't searched any more than that.

 

[Durvelle27]

It's a temp problem bud. It's not the CPU but the actual CPU socket

 

[Ravenas]

It's a temp problem bud. It's not the CPU but the actual CPU socket

I am assuming you are talking about the heat sink? I have reapplied it 3 separate times to no avail.

 

[Frick]

I am assuming you are talking about the heat sink? I have reapplied it 3 separate times to no avail.

He's saying it doesn't report temps corretly.

EDIT: ohh that's what he meant. Scratch my post.

 

[Durvelle27]

I am assuming you are talking about the heat sink? I have reapplied it 3 separate times to no avail.

No the actual back of the motherboard socket. It gets pretty toasty after certain clocks which can cause throttling and freezes.

 

[Ravenas]

He's saying it doesn't report temps corretly.

That may be true... The CPU may be more around 70-80 degrees. The liquid in my water cooler is around 32 degrees during stress tests in OCCT.

No the actual back of the motherboard socket. It gets pretty toasty after certain clocks which can cause throttling and freezes.

The only way to control this would seem to be a larger radiator for my water cooler and maybe additional fans.

 

[Durvelle27]

Also what software are using to read temps as not all support the FX Line

 

[Ravenas]

Also what software are using to read temps as not all support the FX Line

I am using the Asus Suite sensor readings and OCCT displays it real time as well.

 

[Durvelle27]

I am using the Asus Suite sensor readings and OCCT displays it real time as well.

HwInfo64 has always been the most accurate for me

 

[Ravenas]

HwInfo64 has always been the most accurate for me

HwInfo64 is built in to OCCT by default.

 

[Durvelle27]

HwInfo64 is built in to OCCT by default.

What are the motherboard and socket temps

 

[Ravenas]

What are the motherboard and socket temps

The motherboard temp is a constant 30 degrees C and the socket temp is 32 +/- 2 degrees C.

 

[Durvelle27]

That can't be accurate at all. Socket temp will not be that low unless idle

 

[Ravenas]

That can't be accurate at all. Socket temp will not be that low unless idle

Let me check the files again to make sure.

 

[Durvelle27]

Let me check the files again to make sure.

Alright

 

[ThE_MaD_ShOt]

Also to answer your question, yes you need a better cooler. I am running a h100i on my main rig thats clocked at 4.4 and under load it is at 35c. which it's under load 24/7.

 

[Ravenas]

That can't be accurate at all. Socket temp will not be that low unless idle

The CPU package temp is the socket temp correct?

 

[Durvelle27]

The CPU package temp is the socket temp correct?

Yes

Screen shot would help

 

[Ravenas]

 

[OneMoar]

socket temp is meaningless litterally doesn't matter I suspose it could be the VRM OCP kicking in and throttling back causing a bsod
give the vrm heatsink a feel and if you can't keep your finger on it for more then 3 seconds it COULD be getting a bit toasty under a 220W load
Unless your chip is defective it should't be crashing at 60C amd says 62c but you usually have a fair bit more then that before things start going wonkey
id point a fan at the VRM and see what happens with the lack of airflow due to the AIO its probly getting toasty

 

[Steevo]

It actually seems like you may have something obstructing the flow through the block or radiator, the IHS has poor internal contact with the die, and or the pump is dying.


I would take the lines off the block and look inside to make sure nothing is growing and blocking cooling fins, or get a thermocouple and put it on the side of the IHS or cooling block to see if you have a good candidate for removing the IHS.

 

[OneMoar]

It actually seems like you may have something obstructing the flow through the block or radiator, the IHS has poor internal contact with the die, and or the pump is dying.


I would take the lines off the block and look inside to make sure nothing is growing and blocking cooling fins, or get a thermocouple and put it on the side of the IHS or cooling block to see if you have a good candidate for removing the IHS.

wtf are you smoking