I upgraded to an AMD FX-8350 a Corsair H80i liquid cooler back in January. At first, it worked great - the CPU barely broke 60 in a stress test at a mild OC (4.4), and it would idle within a few degrees of the ambient temperature in the room.
After a couple months, I started having serious overheating problems. I didn't realize what was happening at first when I started getting unexpected shutdowns - I thought maybe a bad driver was causing it - until I happened to notice my CPU temperature hit 80 a couple minutes after restarting and trying to play a game. By turning the fans on my H80i radiator block to max speed at all times and underclocking/undervolting as much as I can (can't seem to undervolt below 1.2875 (-0.1) without stability issues, but I'm currently underclocked all the way down to 2.7), and it still idles at 40, hits the 50s while browsing CPU-heavy websites like Facebook, and struggles to stay at *70* while gaming!
The only components that are getting unreasonably hot are the CPU and the northbridge. I've thoroughly cleaned all my fans and the radiator for my H80i, and verified that all my fans are working normally. The voltages coming from my PSU appear to be accurate and stable (although I've only verified this with software, not a multimeter). I've cleaned off and re-applied the thermal paste and reseated the CPU and cooling block multiple times to make sure that wasn't the issue. The thermals definitely look like the CPU is just putting out way too much heat - the CPU temperature spikes up and then goes quickly back down a bit, then jumps around as the CPU load fluctuates (now that I'm underclocked enough that it doesn't just overheat and shut down my system entirely), the cooling block spikes up and stays hot until the CPU is idle again, then quickly cools off, and the northbridge heats up more slowly after the CPU does, then slowly drops down when the CPU stays idle. Basically, it seems like the CPU is putting out too much heat, the problem gets much worse at higher clockspeeds, voltages, and CPU loads, and the liquid cooler is pulling heat off as fast as it can but still can't keep the CPU cool enough at stock settings.
I'm pretty sure it must be the CPU itself, but I've never had to RMA a CPU before so I wanted to make sure I'm not missing some other potential cause. If I do need to RMA it, is AMD going to tell me tough luck because I've changed the CPU multiplier and voided my warranty? Assuming I should RMA it, and at least have a chance of getting a replacement, can anyone recommend a place to buy a temporary replacement CPU that I could return open box if and when my replacement arrives? I don't have my old CPU anymore, and I can't really go for weeks without a usable desktop.
After a couple months, I started having serious overheating problems. I didn't realize what was happening at first when I started getting unexpected shutdowns - I thought maybe a bad driver was causing it - until I happened to notice my CPU temperature hit 80 a couple minutes after restarting and trying to play a game. By turning the fans on my H80i radiator block to max speed at all times and underclocking/undervolting as much as I can (can't seem to undervolt below 1.2875 (-0.1) without stability issues, but I'm currently underclocked all the way down to 2.7), and it still idles at 40, hits the 50s while browsing CPU-heavy websites like Facebook, and struggles to stay at *70* while gaming!
The only components that are getting unreasonably hot are the CPU and the northbridge. I've thoroughly cleaned all my fans and the radiator for my H80i, and verified that all my fans are working normally. The voltages coming from my PSU appear to be accurate and stable (although I've only verified this with software, not a multimeter). I've cleaned off and re-applied the thermal paste and reseated the CPU and cooling block multiple times to make sure that wasn't the issue. The thermals definitely look like the CPU is just putting out way too much heat - the CPU temperature spikes up and then goes quickly back down a bit, then jumps around as the CPU load fluctuates (now that I'm underclocked enough that it doesn't just overheat and shut down my system entirely), the cooling block spikes up and stays hot until the CPU is idle again, then quickly cools off, and the northbridge heats up more slowly after the CPU does, then slowly drops down when the CPU stays idle. Basically, it seems like the CPU is putting out too much heat, the problem gets much worse at higher clockspeeds, voltages, and CPU loads, and the liquid cooler is pulling heat off as fast as it can but still can't keep the CPU cool enough at stock settings.
I'm pretty sure it must be the CPU itself, but I've never had to RMA a CPU before so I wanted to make sure I'm not missing some other potential cause. If I do need to RMA it, is AMD going to tell me tough luck because I've changed the CPU multiplier and voided my warranty? Assuming I should RMA it, and at least have a chance of getting a replacement, can anyone recommend a place to buy a temporary replacement CPU that I could return open box if and when my replacement arrives? I don't have my old CPU anymore, and I can't really go for weeks without a usable desktop.
glad we could help you find the problem. Sounds like something was blocking the coolent flow to your radiator... that or the pump propeller broke (and not the engine)... either way your cpu cooler was certainly behaving like a water cooler with no waterflow.
Facebook
Twitter
Instagram