rranney
02-02-2010, 08:36 PM
Problem:
Recent renders from Maya 2009, after several minutes of using 100% CPU load on quad-core, results in a thermal shutdown. Seems that the CPU is overheating suddenly.
I have attached a series of screenshots and photos to help as visual aids.
System Specs:
Custom built desktop machine (nothing is over-clocked) 2yrs old
OS: Windows XP SP3
Processor: Intel Core 2 Quad Q6600 GO @ 2.4Ghz
Heatsink: Intel stock Core 2 Duo Heatsink (http://www.frostytech.com/articleview.cfm?articleID=2132)
Motherboard: EVGA nVidia nForce 680i SLI [driver version 15.46]
Graphics Card: nVidia GeForce 8800 GTS 640MB [driver version 196.21]
Memory: GSKILL 4GB (2x2GB) 240pin DDR2 SDRAM
Power Supply: Antec Neo Power 500W ATX12V
Hard Drives: (no RAID setup)
1) WD Raptor 150GB 10,000rpm (SATA)...OS installed here
2) WD Caviar 320GB 7,200rpm (SATA)
3) WD Caviar 320GB 7,200rpm (SATA)
4) WD Caviar 640GB 7,200rpm (SATA) (recently added, but problem occurred before)
Case: ANTEC P182
Other:
ASUS CD/DVD R/W drive
GIGAWARE 4port USB hub
NETGEAR WG1311v3 Wireless PCI Adaptor
Two Samsung LCD 17in monitors
M-AUDIO Fast Track Pro USB device (I pipe my sound through this to use Pro Tools)
Notes:
All drives are deframented, none approaching full capacity. Ambient room temp a dry 22C. Tower sits on desktop, not crowded or choking on carpet floor. Case has three 120mm fans.
What I've tried:
I recently updated the BIOS and chipset drivers for my motherboard, so I'm suspicious whether this changed how my system is monitoring my temperatures. I read a post in a different forum that someone fixed a similar problem by rolling back their mobo updates. This is the first time I've ran into this problem, so I starting monitoring my temps to troubleshoot and wolla, hot CPU. Though I don't know what my temps were for the last two years or prior to the motherboard updates, I never knew about temp monitoring or had a reason to check.
Using temp monitors like RealTemp, CoreTemp and Nvidia's monitor, I have tracked my CPU temp at 60C at idle. From my research, I hear this is supposed to be in the 30s-40s. Upon startup in the morning (after a night of cooling), it's around 50, but it only takes a minute or two for it to climb to 60C at idle. Upon thermal shutdown, CoreTemp displays temps beyond 100C (Tjmax).
I first gave my entire case a thorough cleaning: every fan (CPU, GPU, 3 case fans, mobo fan), and used compressed air to get all the dust out from nooks and crannies, and the surfaces of components. I was expecting some sort of CPU temp difference (heatsink was really dirty), but I noted no change in the temp monitors.
Each temp monitor displays two of my cores 10C hotter than the other two. Not sure why that is...
While the CPU temp redlines in the 90s, the heatsink aluminum fins don't feel hot to the touch, only slightly warm.
While troubleshooting some BSOD earlier this month, I set my BIOS to default settings several times. I just discovered that this turns off thermal regulation of the Intel chip (TM1 and TM2). So I know I have successfully operated without the TM1 or TM2 CPU thermal regulation before. When I turn on TM1 (thermal throttling by inserting idle cycles), Maya is able to render without thermal shutdown, but the CPU temp stays in the 90s celsius. So regardless of finding a way to get renders out, my CPU is still very hot during idle and under load if the sensor temperatures are correct. This is even with an open case.
My questions:
1) Could my BIOS update account for different temperature monitoring / change in the sensors? Could resetting the BIOS (taking the battery out, putting back in) change anything? I'd rather not touch hardware, thermal paste, and heatsink upgrades if my sensors are simply outta whack...
2) The only thing I can think of left to try is buying some new thermal paste (Arctic Silver 5) and reapply my heatsink fan. Although I've never had this issue before, why do I suddenly experience overheating?
3) I believe my GPU temp is satisfactory (around 60C), but what are motherboard temps supposed to be? If the temps are all inflated, it could be a sign of faulty sensors (or cause for alarm).
Thanks everyone for any advice! I know this post was long, but I read the sticky about being as descriptive as possible...
Recent renders from Maya 2009, after several minutes of using 100% CPU load on quad-core, results in a thermal shutdown. Seems that the CPU is overheating suddenly.
I have attached a series of screenshots and photos to help as visual aids.
System Specs:
Custom built desktop machine (nothing is over-clocked) 2yrs old
OS: Windows XP SP3
Processor: Intel Core 2 Quad Q6600 GO @ 2.4Ghz
Heatsink: Intel stock Core 2 Duo Heatsink (http://www.frostytech.com/articleview.cfm?articleID=2132)
Motherboard: EVGA nVidia nForce 680i SLI [driver version 15.46]
Graphics Card: nVidia GeForce 8800 GTS 640MB [driver version 196.21]
Memory: GSKILL 4GB (2x2GB) 240pin DDR2 SDRAM
Power Supply: Antec Neo Power 500W ATX12V
Hard Drives: (no RAID setup)
1) WD Raptor 150GB 10,000rpm (SATA)...OS installed here
2) WD Caviar 320GB 7,200rpm (SATA)
3) WD Caviar 320GB 7,200rpm (SATA)
4) WD Caviar 640GB 7,200rpm (SATA) (recently added, but problem occurred before)
Case: ANTEC P182
Other:
ASUS CD/DVD R/W drive
GIGAWARE 4port USB hub
NETGEAR WG1311v3 Wireless PCI Adaptor
Two Samsung LCD 17in monitors
M-AUDIO Fast Track Pro USB device (I pipe my sound through this to use Pro Tools)
Notes:
All drives are deframented, none approaching full capacity. Ambient room temp a dry 22C. Tower sits on desktop, not crowded or choking on carpet floor. Case has three 120mm fans.
What I've tried:
I recently updated the BIOS and chipset drivers for my motherboard, so I'm suspicious whether this changed how my system is monitoring my temperatures. I read a post in a different forum that someone fixed a similar problem by rolling back their mobo updates. This is the first time I've ran into this problem, so I starting monitoring my temps to troubleshoot and wolla, hot CPU. Though I don't know what my temps were for the last two years or prior to the motherboard updates, I never knew about temp monitoring or had a reason to check.
Using temp monitors like RealTemp, CoreTemp and Nvidia's monitor, I have tracked my CPU temp at 60C at idle. From my research, I hear this is supposed to be in the 30s-40s. Upon startup in the morning (after a night of cooling), it's around 50, but it only takes a minute or two for it to climb to 60C at idle. Upon thermal shutdown, CoreTemp displays temps beyond 100C (Tjmax).
I first gave my entire case a thorough cleaning: every fan (CPU, GPU, 3 case fans, mobo fan), and used compressed air to get all the dust out from nooks and crannies, and the surfaces of components. I was expecting some sort of CPU temp difference (heatsink was really dirty), but I noted no change in the temp monitors.
Each temp monitor displays two of my cores 10C hotter than the other two. Not sure why that is...
While the CPU temp redlines in the 90s, the heatsink aluminum fins don't feel hot to the touch, only slightly warm.
While troubleshooting some BSOD earlier this month, I set my BIOS to default settings several times. I just discovered that this turns off thermal regulation of the Intel chip (TM1 and TM2). So I know I have successfully operated without the TM1 or TM2 CPU thermal regulation before. When I turn on TM1 (thermal throttling by inserting idle cycles), Maya is able to render without thermal shutdown, but the CPU temp stays in the 90s celsius. So regardless of finding a way to get renders out, my CPU is still very hot during idle and under load if the sensor temperatures are correct. This is even with an open case.
My questions:
1) Could my BIOS update account for different temperature monitoring / change in the sensors? Could resetting the BIOS (taking the battery out, putting back in) change anything? I'd rather not touch hardware, thermal paste, and heatsink upgrades if my sensors are simply outta whack...
2) The only thing I can think of left to try is buying some new thermal paste (Arctic Silver 5) and reapply my heatsink fan. Although I've never had this issue before, why do I suddenly experience overheating?
3) I believe my GPU temp is satisfactory (around 60C), but what are motherboard temps supposed to be? If the temps are all inflated, it could be a sign of faulty sensors (or cause for alarm).
Thanks everyone for any advice! I know this post was long, but I read the sticky about being as descriptive as possible...
