It's all much quieter now too which is presumably because the GPU speeds everything up and doesn't have long enough to get too hot. Now that I have ticked the appropriate box in Preferences, following HO72's advice, every action is much faster (night and day difference) and the Task Manager confirms it: for the now brief times the app is working, the CPU and GPU percentages both rise with the GPU briefly going above 80%. It works well but coming back to LR it is a TIFF in a DNG wrapper. Also, Topaz is still quite slow! Takes 15 seconds to run AI Clear on a 26MP image.Īh yes, that's very interesting. p.1 9 p.1 9 How do you use Topaz DeNoise AI products You can drag a file directly from LR or from your OS folder structure into DeNoise which allows you to edit as a RAW. So it is definitely using my Graphics card However, it doesn't cause my fans to ramp up unless I've done a huge set. On mine, with an Radeon RX570 and similar speed CPU to yours, I can see it hit 80-90% usage when running AI tasks in Topaz. If you GPU (Graphics card) is being used, you will see the GPU usage percentage climb.
While making an edit, watch the columns for CPU and GPU. Find the Topaz application under the processes tab. While running Topaz, open up task manager (right click task bar, Task Manager).
#Topaz denoise ai fails to run full#
If the GPU cooler is not working that could be a reason the CPU fan is going full bore. Like immediately, before doing anything else. The OP needs to use a utility, eg HW Monitor, to check the temps of the GPU if the fans are not spinning. The former accounts for the speed of booting a computer from solid state storage but bottlenecks in the system tend to limit the perceptible differences in boot times between solid state media types. The I/O speed benchmarks generated in something like Crystal Mark are not consistently perceived in real world use except for large sustained reads and writes. For files under 100mb the difference in times writing over SATA or PCIe/M2 bus to any internal drive is fractions of a second.As files get larger perceptible differences in write times, writes always being slower than reads, are more apparent as is the impact of using an external bus. If the OP is storing a large PSD or TIFF file as the final product it takes the time to write to storage that it takes. I am going to bet that the GPU makes little if any perceptible difference. In that case a sense of rendering time with and without the new GPU can be inferred in how long it takes to get the finished image from Topaz back to Photoshop/LR. I will presume the OP is using Topaz as a plug-in. The two can be difficult if not impossible to suss out. Based on what the OP writes I am guessing she is conflating I/0 with rendering times.
0 kommentar(er)
