Video AI v3.1.X - User Benchmarking Results

ole.solev · March 15, 2023, 3:19pm

DDR4, GPU driver 531.29 Game Ready.

Topaz Video AI  v3.1.9
System Information
OS: Windows v11.2009
CPU: 12th Gen Intel(R) Core(TM) i5-12600K  31.779 GB
GPU: NVIDIA GeForce RTX 3060  11.845 GB
GPU: Intel(R) UHD Graphics 770  0.125 GB
Processing Settings: device: 0 vram: 1 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis		1X: 9.38 fps 	2X: 5.66 fps 	4X: 1.67 fps 	
Proteus		1X: 8.05 fps 	2X: 5.3 fps 	4X: 1.61 fps 	
Gaia		1X: 3.11 fps 	2X: 2.14 fps 	4X: 1.38 fps 	
4X Slowmo		Apollo: 8.59 fps 	Chronos: 7.04 fps 	Chronos Fast: 10.65 fps

Topaz Video AI  v3.1.9
System Information
OS: Windows v11.2009
CPU: 12th Gen Intel(R) Core(TM) i5-12600K  31.779 GB
GPU: NVIDIA GeForce RTX 3060  11.845 GB
GPU: Intel(R) UHD Graphics 770  0.125 GB
Processing Settings: device: 0 vram: 1 instances: 1
Input Resolution: 720x480
Benchmark Results
Artemis		1X: 53.06 fps 	2X: 33.99 fps 	4X: 9.95 fps 	
Proteus		1X: 51.48 fps 	2X: 34.23 fps 	4X: 11.3 fps 	
Gaia		1X: 18.1 fps 	2X: 12.64 fps 	4X: 8.36 fps 	
4X Slowmo		Apollo: 51.24 fps 	Chronos: 38.51 fps 	Chronos Fast: 51.3 fps

ole.solev · March 15, 2023, 3:25pm

will 3.2.* follow 3.1.9?
will 3.2.* be a major release?

It seems that you really need to show the version of the GPU driver in the report.

lemans99 · March 15, 2023, 3:27pm

yeah - same GPU different CPU - just double the performance …

lemans99 · March 15, 2023, 3:35pm

I think no RAM shortage even at 16GB … But guess DDR5 doing good against DDR3 Still wonder how better is Win11 then Win10 …

lemans99 · March 15, 2023, 3:42pm

Well i am not sure if DDR4 is OK … Looking at the two 4090 performance above . The one paired with 13900KF gives twice better performance then the one with Ryzen 5950X …
My driver is v531 .

ole.solev
p.s Whats the point to delete your post after i did replay ?!

toddsnc · March 15, 2023, 4:02pm

Mine is the 4090 with the i9-13900KF. It’s running DDR5-5600, XMP-II. No components are overclocked. Samsung NVMe 980 Pro 2TB. MSI Gaming GeForce RTX 4090 24GB GDDR6X PCI Express 4.0 GPU.

mallguren · March 15, 2023, 4:50pm

Here’s mine at 3840x2160:

Topaz Video AI  v3.1.9
System Information
OS: Windows v10.2009
CPU: AMD Ryzen Threadripper 3960X 24-Core Processor   127.87 GB
GPU: NVIDIA GeForce RTX 2080 Ti  10.782 GB
Processing Settings: device: 0 vram: 0.9 instances: 1
Input Resolution: 3840x2160
Benchmark Results
Artemis		1X: 2.95 fps 	2X: 1.43 fps 	4X: 0.42 fps 	
Proteus		1X: 2.49 fps 	2X: 1.33 fps 	4X: 0.43 fps 	
Gaia		1X: 1.05 fps 	2X: 0.68 fps 	4X: 0.42 fps 	
4X Slowmo		Apollo: 3.91 fps 	Chronos: 0.94 fps 	Chronos Fast: 3.97 fps

ArtndFun · March 15, 2023, 5:33pm

Nice stuff this feature.
I recommend copy pasting if someone wants to search for a specific model.

Topaz Video AI  v3.1.9
System Information
OS: Windows v10.2009
CPU: AMD Ryzen 7 5700G with Radeon Graphics           15.342 GB
GPU: NVIDIA GeForce RTX 2060  5.8281 GB
GPU: AMD Radeon(TM) Graphics  0.48415 GB
Processing Settings: device: 0 vram: 1 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis		1X: 7.63 fps 	2X: 4.46 fps 	4X: 1.25 fps 	
Proteus		1X: 6.66 fps 	2X: 4.19 fps 	4X: 1.24 fps 	
Gaia		1X: 2.56 fps 	2X: 1.67 fps 	4X: 1.04 fps 	
4X Slowmo		Apollo: 6.84 fps 	Chronos: 2.33 fps 	Chronos Fast: 8.41 fps

preston.werner · March 15, 2023, 7:01pm

Correct, the ‘major’ versions will be as follows;

3.2.X
3.3.X
3.4.X

et cetera

vargolsoft · March 15, 2023, 9:11pm

I thought A base M1 set of results might come in handy.


Topaz Video AI v3.1.9
System Information
OS: Mac v13.0201
CPU: Apple M1 8 GB
GPU: Apple M1 5.3333 GB
Processing Settings: device: 0 vram: 1 instances: 1

Input Resolution: 1920x1080

Benchmark Results
Artemis 1X: 3.69 fps 2X: 2.5 fps 4X: 0.87 fps
Proteus 1X: 3.73 fps 2X: 2.48 fps 4X: 0.41 fps
Gaia 1X: 1.22 fps 2X: 0.9 fps 4X: 0.7 fps
4X Slowmo Apollo: 4.3 fps Chronos: 1.22 fps Chronos Fast: 1.88 fps

ForSerious · March 15, 2023, 9:24pm

I had to turn the VRAM setting down to 11 to get it to run all the benchmarks:

Topaz Video AI  v3.1.9
System Information
OS: Windows v10.2009
CPU: Intel(R) Core(TM) i5-4690K CPU @ 3.50GHz  7.953 GB
GPU: NVIDIA GeForce GTX 1060 3GB  2.9082 GB
Processing Settings: device: 0 vram: 0.11 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis		1X: 0.54 fps 	2X: 0.39 fps 	4X: 0.16 fps 	
Proteus		1X: 0.53 fps 	2X: 0.39 fps 	4X: 0.17 fps 	
Gaia		1X: 0.18 fps 	2X: 0.12 fps 	4X: 0.06 fps 	
4X Slowmo		Apollo: 0.86 fps 	Chronos: 0.36 fps 	Chronos Fast: 0.55 fps

ForSerious · March 15, 2023, 9:29pm

I think your CPU just doesn’t have enough cores. If you can increase your RAM speed, that should speed it up some.

Topaz_Usr · March 15, 2023, 10:21pm

Topaz Video AI  v3.1.9
System Information
OS: Windows v10.2009
CPU: AMD Ryzen 9 5900X 12-Core Processor              127.9 GB
GPU: NVIDIA GeForce RTX 3090  23.77 GB
Processing Settings: device: 0 vram: 0.8 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis		1X: 19.65 fps 	2X: 7.87 fps 	4X: 1.95 fps 	
Proteus		1X: 13.98 fps 	2X: 6.13 fps 	4X: 1.93 fps 	
Gaia		1X: 7.67 fps 	2X: 4.81 fps 	4X: 2.63 fps 	
4X Slowmo		Apollo: 16.48 fps 	Chronos: 15.93 fps 	Chronos Fast: 23.25 fps

lpenderg · March 15, 2023, 10:22pm

Topaz Video AI  v3.1.9
System Information
OS: Windows v10.2009
CPU: Intel(R) Core(TM) i7-4770K CPU @ 3.50GHz  31.877 GB
GPU: NVIDIA GeForce RTX 2080  7.8145 GB
GPU: Intel(R) HD Graphics 4600  0.10986 GB
Processing Settings: device: -2 vram: 0.9 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis		1X: 8.73 fps 	  2X: 3.34 fps 	4X: 0.91 fps 	
Proteus		1X: 6.83 fps 	  2X: 3.1 fps 	        4X: 0.91 fps 	
Gaia		1X: 3.46 fps 	  2X: 2.36 fps 	4X: 1.19 fps 	
4X Slowmo		Apollo: 7.48 fps 	Chronos: 3.02 fps 	Chronos Fast: 11.44 fps

david-1264 · March 15, 2023, 10:51pm

These are my results. I’ve added 2 results for the 1080p test to show the different results.
I think it might be better if it processed a longer video during the test, at least for the 1080p test, that way it could give more accurate average results, since it gives results to 2 dp but some of them are quite different if you run the test again eg. 16.14 fps on one test and 16.69 fps for Artemis 1x on the next test.

Also the test shows “vram: 0.9” - I assume that’s the “max memory usage percentage” - but it doesn’t mention that it’s vram in the preferences. Maybe it should if it is and maybe the test results could be a bit clearer about that.

Topaz Video AI  v3.1.9
System Information
OS: Windows v10.2009
CPU: AMD Ryzen Threadripper 3960X 24-Core Processor   255.88 GB
GPU: NVIDIA GeForce RTX 3090  23.77 GB
Processing Settings: device: 0 vram: 0.9 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis		1X: 16.14 fps 	2X: 5.84 fps 	4X: 1.89 fps 	
Proteus		1X: 10.85 fps 	2X: 4.95 fps 	4X: 1.79 fps 	
Gaia		1X: 7.25 fps 	2X: 4.74 fps 	4X: 2.37 fps 	
4X Slowmo		Apollo: 17.32 fps 	Chronos: 15.35 fps 	Chronos Fast: 21.28 fps

Topaz Video AI  v3.1.9
System Information
OS: Windows v10.2009
CPU: AMD Ryzen Threadripper 3960X 24-Core Processor   255.88 GB
GPU: NVIDIA GeForce RTX 3090  23.77 GB
Processing Settings: device: 0 vram: 0.9 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis		1X: 16.69 fps 	2X: 5.83 fps 	4X: 1.85 fps 	
Proteus		1X: 11.19 fps 	2X: 5.49 fps 	4X: 1.85 fps 	
Gaia		1X: 7.3 fps 	2X: 4.62 fps 	4X: 2.43 fps 	
4X Slowmo		Apollo: 17.79 fps 	Chronos: 15.51 fps 	Chronos Fast: 20.45 fps

Topaz Video AI  v3.1.9
System Information
OS: Windows v10.2009
CPU: AMD Ryzen Threadripper 3960X 24-Core Processor   255.88 GB
GPU: NVIDIA GeForce RTX 3090  23.77 GB
Processing Settings: device: 0 vram: 0.9 instances: 1
Input Resolution: 3840x2160
Benchmark Results
Artemis		1X: 3.63 fps 	2X: 1.33 fps 	4X: 0.35 fps 	
Proteus		1X: 2.48 fps 	2X: 1.16 fps 	4X: 0.38 fps 	
Gaia		1X: 1.53 fps 	2X: 0.99 fps 	4X: 0.53 fps 	
4X Slowmo		Apollo: 4.33 fps 	Chronos: 3.35 fps 	Chronos Fast: 5.63 fps

ForSerious · March 16, 2023, 12:28am

AMD Ryzen CPUs run faster with faster RAM. Not sure about Intel CPUs.
My Ryzen has 12 cores and seems to load about 10 of them pretty heavy while processing with TVAI.
I have not tried disabling cores to see if that slows things down, but I did try running two instances at the same time. The speed gain, looked pretty close to what I would expect if I were to somehow force it to use all 12 cores with one instance.
Another little quirk of, at least my CPU, is that it boosts frequencies much more if not all the cores are loaded.

WALTER_BLACK · March 16, 2023, 4:04am

Topaz Video AI  v3.1.9
System Information
OS: Windows v11.2009
CPU: Intel(R) Core(TM) i7-8700K CPU @ 3.70GHz  31.946 GB
GPU: Intel(R) Arc(TM) A770 Graphics  15.875 GB
Processing Settings: device: 0 vram: 1 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis		1X: 9.92 fps 	2X: 5.15 fps 	4X: 1.53 fps 	
Proteus		1X: 8.79 fps 	2X: 4.21 fps 	4X: 1.28 fps 	
Gaia		1X: 5.46 fps 	2X: 3.31 fps 	4X: 2.77 fps 	
4X Slowmo		Apollo: 8.17 fps 	Chronos: 2.85 fps 	Chronos Fast: 4.91 fps

is this results normal for A770 16G ?

ole.solev · March 16, 2023, 6:40am

Does the encoding speed affect the benchmark?

lemans99 · March 16, 2023, 7:27am

I think Arc 770 performs perfect for its price ! But so far No anyone with AMD GPU !?? Really ?? This is so strange …

tong.beng_hwa · March 16, 2023, 9:22am

Topaz Video AI  v3.1.9
System Information
OS: Windows v11.2009
CPU: AMD Ryzen 9 5950X 16-Core Processor              31.924 GB
GPU: NVIDIA GeForce RTX 4090  23.59 GB
Processing Settings: device: 0 vram: 1 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis		1X: 23.31 fps 	2X: 9.16 fps 	4X: 2.16 fps 	
Proteus		1X: 16.07 fps 	2X: 7.29 fps 	4X: 2.24 fps 	
Gaia		1X: 13.86 fps 	2X: 8.78 fps 	4X: 2.91 fps 	
4X Slowmo		Apollo: 18.46 fps 	Chronos: 26.25 fps 	Chronos Fast: 28.55 fps