Associate
- Joined
- 4 May 2009
- Posts
- 188
- Location
- Edinburgh, UK
Hi folks, I'm after a bit of advice, please.
I have a 2080ti in my system with a Ryzen 3700X. It's watercooled (CPU & GPU) with 2x 360 rads. Water temp is 33-35C, with CPU temp never getting higher than 65C, after a long time under load.
I'm using Aida64 for monitoring, so I have a good amount of information about my system. What I've discovered is that if my GPU runs at around 150-200W, the system can run basically forever. And some games (e.g. Elite Dangerous) never stress the card harder than that.
However, other games (e.g. Cyberpunk) will see the card pushing 250-250W. At this higher power consumption I see an increasing gap between the GPU Diode Temp and the GPU Hotspot temp. The diode stays at around 50C but the hotspot climbs and climbs until the card crashes at around 86C, necessitating a hard reset.
I've done a little bit of reading about this, and it looks like I can cure this by re-pasting and re-seating my block. It's about 18 months since I last did that.
Do you agree with that advice? Or do you think I'm wasting my time?
Is my observation about it being power-related a red herring? Aside from using Afterburner to reduce the power limit, what changes would allow me to restrain the Hotspot temp without losing too much performance?
I have a 2080ti in my system with a Ryzen 3700X. It's watercooled (CPU & GPU) with 2x 360 rads. Water temp is 33-35C, with CPU temp never getting higher than 65C, after a long time under load.
I'm using Aida64 for monitoring, so I have a good amount of information about my system. What I've discovered is that if my GPU runs at around 150-200W, the system can run basically forever. And some games (e.g. Elite Dangerous) never stress the card harder than that.
However, other games (e.g. Cyberpunk) will see the card pushing 250-250W. At this higher power consumption I see an increasing gap between the GPU Diode Temp and the GPU Hotspot temp. The diode stays at around 50C but the hotspot climbs and climbs until the card crashes at around 86C, necessitating a hard reset.
I've done a little bit of reading about this, and it looks like I can cure this by re-pasting and re-seating my block. It's about 18 months since I last did that.
Do you agree with that advice? Or do you think I'm wasting my time?
Is my observation about it being power-related a red herring? Aside from using Afterburner to reduce the power limit, what changes would allow me to restrain the Hotspot temp without losing too much performance?
