Hello,
I recently decided to 'refresh' my now almost decade old secondary desktop (by squeezing out some more performance with help of OC until CPU/GPU prices drop to something reasonable).
CPU is AMD FX-8320 in Asus M5A97 evo 2.0 motherboard (I added my brothers old Radeon R9 270X and fresh samsung 860 ssd). Cooler is old AC freezer xtreme (for 140W TDP; I think they don't make them anymore and it shows how old this system is
)
I tried running the system at stock voltages at 4.0 GHz and it was perfectly stable albeit CPU socket temperature kept climbing to over ~80C, package and core temps were generally below 60-65C after pelting it for 6 hours with prime95.
I decided to run it at 3.9GHz which gave me barely comfortable 77C on socket after few hours of grinding primes.
So, a modest overclock due to insufficient cooling (I did not even touch the voltages!). CPU itself is OK, it's the socket that gets too hot.
Is there any way to lower CPU socket temperatures? Chip can probably take higher OC, but I will melt the socket at that rate
Now here comes the weird part:
Some time after (few hours?) my screen turned green, but not like GSOD - everything seems to work perfectly fine apart from green tint on screen and it was persistent across reboots.
I tried second port on my graphics card and it was fine for a while, but it too turned permanently green after ~30 minutes. Now both of them emit green picture (they are DVI ports, HDMI is fine so far on the second screen).
Monitor problem is out of the question because I tested it on another computer (with cable) and it works fine.
I suspected GPU, but I booted into linux and green tint is gone - and it is back every time I reboot to Win10. Unplugging and re-plugging monitor makes no difference at all.
I did factory reset of Radeon Software settings and it temporarily fixed the issue, but screen turns back to green after few hours of prime95 (with absolutely ZERO errors each time)
I feel that OCing has something to do with this, but I'm stumped. I reverted settings back to stock and ran prime95 for 24h with no green in sight.
Has anyone encountered anything similar to this?
Edit: GPU was sitting idle at 36C all the time
I recently decided to 'refresh' my now almost decade old secondary desktop (by squeezing out some more performance with help of OC until CPU/GPU prices drop to something reasonable).
CPU is AMD FX-8320 in Asus M5A97 evo 2.0 motherboard (I added my brothers old Radeon R9 270X and fresh samsung 860 ssd). Cooler is old AC freezer xtreme (for 140W TDP; I think they don't make them anymore and it shows how old this system is

I tried running the system at stock voltages at 4.0 GHz and it was perfectly stable albeit CPU socket temperature kept climbing to over ~80C, package and core temps were generally below 60-65C after pelting it for 6 hours with prime95.
I decided to run it at 3.9GHz which gave me barely comfortable 77C on socket after few hours of grinding primes.
So, a modest overclock due to insufficient cooling (I did not even touch the voltages!). CPU itself is OK, it's the socket that gets too hot.
Is there any way to lower CPU socket temperatures? Chip can probably take higher OC, but I will melt the socket at that rate
Now here comes the weird part:
Some time after (few hours?) my screen turned green, but not like GSOD - everything seems to work perfectly fine apart from green tint on screen and it was persistent across reboots.
I tried second port on my graphics card and it was fine for a while, but it too turned permanently green after ~30 minutes. Now both of them emit green picture (they are DVI ports, HDMI is fine so far on the second screen).
Monitor problem is out of the question because I tested it on another computer (with cable) and it works fine.
I suspected GPU, but I booted into linux and green tint is gone - and it is back every time I reboot to Win10. Unplugging and re-plugging monitor makes no difference at all.
I did factory reset of Radeon Software settings and it temporarily fixed the issue, but screen turns back to green after few hours of prime95 (with absolutely ZERO errors each time)
I feel that OCing has something to do with this, but I'm stumped. I reverted settings back to stock and ran prime95 for 24h with no green in sight.
Has anyone encountered anything similar to this?
Edit: GPU was sitting idle at 36C all the time
Last edited: