I have been using Topaz Video successfully for a few weeks. Then suddenly in the middle of one of the projects, the frame rate dropped to zero and effectively stopped processing. I deleted the video and Topaz did not proceed to the next queued project. I checked and my Nvidia RTX5060 driver is current. I haven’t tried to uninstall and reinstall Topaz, but wondered if someone else has experienced this and can offer a solution. Thanks!
This actually sounds like a hardware issue rather than a Topaz bug, specifically thermal throttling.
I suspect you’re experiencing the ‘Pump-out effect’ on your thermal paste. After weeks of heavy rendering (which Topaz excels at), the repeated thermal expansion and contraction can push the paste out from between the GPU die and the heatsink. When this happens, the temperature spikes instantly, and the RTX 5060 throttles its clock speed to zero to protect itself, causing the frame rate to stall.
Before reinstalling everything, check your GPU temperatures with a tool like HWMonitor or GPU-Z while Topaz is running. If you see it hitting 90-100°C before the crash, you likely need a repaste (ideally with a high-viscosity paste or a PTM7950 pad to avoid future pump-out).
Thank you for this information. I purchased a desktop system this past November that was specifically designed for such processing. This does not mean that your diagnosis is incorrect, it’s just that I hoped to avoid such an issue. My workstation is in a Ryzen oversize cabinet that has a dozen cooling fans in it and the CPU is liquid cooled and cycles between 0 and 30% of capacity but generally does not exceed 20% during processing of the videos. I realize that even operating at this capacity can generate a lot of heat as felt from the exhaust fans. In downloading and looking at GPU-z in the middle of rendering, the CPU temp measures between 56-63C and so it seems that the cooling system is working pretty well.
I was able to get Topaz to work again by deleting all the queued projects and beginning anew. And I’ll do as you advise and monitor the CPU temperature. Thanks again for your time and information as this may also help others, too!
I have a 4090 and I must admit that processes lasting several days at full load with Starlight have definitely caused a pump-out effect in my case, as the difference between the GPU temperature and the hotspot temperature keeps increasing.
At least now you’re aware of the phenomenon.
Personally, I recommend running the GPU at a maximum of 75°C and limiting this with Afterburner to avoid this known and not entirely isolated issue.
Since I’m going to switch the GPU to water cooling, I’ll see if this hypothesis is confirmed when I take it apart. I’ll post it here.
I very much appreciate your time and information. I’m a believer in preventative maintenance and, based on your suggestion, will monitor the CPU temperature to avoid the pitfalls of too high of heat. And I send along my best wishes for your success with video rendering!
