How Does the Radeon 9070 XT Perform in Topaz Compared to the 7900 XTX?

Topaz Video  v1.0.1
System Information
OS: Windows v11.24
CPU: Intel(R) Core(TM) i5-14500  95.755 GB
GPU: AMD Radeon RX 9070 XT  15.922 GB
Processing Settings
device: 0 vram: 1 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis		1X: 	27.86 fps 	2X: 	12.97 fps 	4X: 	03.78 fps 	
Iris		1X: 	35.38 fps 	2X: 	15.69 fps 	4X: 	04.50 fps 	
Proteus		1X: 	33.61 fps 	2X: 	16.23 fps 	4X: 	04.63 fps 	
Gaia		1X: 	18.23 fps 	2X: 	10.93 fps 	4X: 	04.39 fps 	
Nyx		1X: 	14.79 fps 	2X: 	11.49 fps 	
Nyx Fast		1X: 	28.91 fps 	
Nyx XL		1X: 	32.46 fps 	
Rhea		4X: 	03.61 fps 	
RXL		4X: 	03.55 fps 	
Hyperion HDR		1X: 	24.23 fps 	
4X Slowmo		Apollo: 	27.96 fps 	APFast: 	56.87 fps 	Chronos: 	20.80 fps 	CHFast: 	30.30 fps 	
16X Slowmo		Aion: 	46.15 fps 	

Topaz Video  v1.0.1
System Information
OS: Windows v11.24
CPU: Intel(R) Core(TM) i5-14500  95.755 GB
GPU: AMD Radeon RX 9070 XT  15.922 GB
Processing Settings
device: 0 vram: 1 instances: 1
Input Resolution: 1280x720
Benchmark Results
Artemis		1X: 	56.54 fps 	2X: 	32.50 fps 	4X: 	08.93 fps 	
Iris		1X: 	62.56 fps 	2X: 	38.41 fps 	4X: 	09.88 fps 	
Proteus		1X: 	65.01 fps 	2X: 	38.69 fps 	4X: 	10.85 fps 	
Gaia		1X: 	36.56 fps 	2X: 	24.74 fps 	4X: 	09.99 fps 	
Nyx		1X: 	22.09 fps 	2X: 	21.29 fps 	
Nyx Fast		1X: 	55.51 fps 	
Nyx XL		1X: 	72.72 fps 	
Rhea		4X: 	08.32 fps 	
RXL		4X: 	07.99 fps 	
Hyperion HDR		1X: 	40.55 fps 	
4X Slowmo		Apollo: 	77.42 fps 	APFast: 	156.77 fps 	Chronos: 	40.79 fps 	CHFast: 	63.47 fps 	
16X Slowmo		Aion: 	68.22 fps 	

I’m noticing a problem where it benchmark quite well. However actual upscale export is really poor.

I have tried upscaling multiple 720p to 1080p on Proteus and I’m only getting around 10-15fps.

Is anyone else having the same issue?

Have you tried a round number upscale? They usually process faster a sharper in my experience. So, try to go from 720 to 1440 instead of 1080.

What’s the output format? That can make a difference.

1 Like

Yes I have tried 2x instead of 1080p as well as trying different output codec. They are all the same.

Currently I am exporting it to FFV1 and then using handbrake to convert it to x265

That is a good way to do it, and shouldn’t be that slow.

The only other thing I can think of is your CPU doesn’t have a lot of fast cores. That might be why, but it’s just a guess.

I think too that the CPU is the Bottleneck.

I also think the problem is my CPU. But I wonder why it performs well in the benchmark.

To my knowledge, the CPU is busy writing the file and therefore cannot supply the GPU.

Compressed and non-hardware-accelerated file formats in particular use a lot of CPU power.