Based on this, it sounds like they made Zen 3 work on older boards via BIOS updates.
Mine is a B350 motherboard.
After receiving criticism from many people, they finally released a BIOS update in 2022 (2 years after the release of 5800x). 
The other reason I switched to Intel was that the AM5 x670 motherboard was quite expensive last year, costing over $300. Meanwhile, Intel Z690 motherboards had a significant discount (roughly half the price) after the release of the Z790 motherboard.
1 Like
Do you have an ASUS board? They have a setting that skips the memory retraining - mine boots up very quickly.
MSI. I have not looked for that. I really shouldnât mind. I restart that computer once every few months.
Topaz Video AI v4.0.2
System Information
OS: Windows v10.22
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: 0 vram: 1 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis 1X: 11.73 fps 2X: 06.81 fps 4X: 01.88 fps
Iris 1X: 11.64 fps 2X: 06.40 fps 4X: 01.72 fps
Proteus 1X: 11.12 fps 2X: 06.22 fps 4X: 02.22 fps
Gaia 1X: 03.93 fps 2X: 02.61 fps 4X: 01.59 fps
Nyx 1X: 04.73 fps 2X: 04.10 fps
4X Slowmo Apollo: 05.03 fps APFast: 31.14 fps Chronos: 09.00 fps CHFast: 12.82 fps
These are pretty abysmal numbers for a $4k computer. Is this software just completely hobbled on Apple hardware?
For Iris, yes.
And as written above I donât understand why it shouldnât be possible to load the old fast model on the e.g. M2 Ultra and the extremely slow one with the artifact-fix only on those models where the error occurred (AFAIK M1 Max).
Or maybe even just provide the old fast Iris V1 and V2 models as a separate download for users to try.
Topaz Video AI v4.0.2
System Information
OS: Mac v14
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: 03.60 fps 2X: 02.33 fps 4X: 00.27 fps
Iris 1X: 02.86 fps 2X: 00.32 fps 4X: ERR fps
Proteus 1X: 03.51 fps 2X: 02.29 fps 4X: ERR fps
Gaia 1X: 01.19 fps 2X: 00.83 fps 4X: 00.28 fps
Nyx 1X: 01.03 fps 2X: 00.92 fps
4X Slowmo Apollo: 03.82 fps APFast: 15.03 fps Chronos: 01.20 fps CHFast: 01.74 fps
Topaz Video AI v4.0.2
System Information
OS: Windows v10.22
CPU: AMD Ryzen 5 3600 6-Core Processor 31.951 GB
GPU: NVIDIA GeForce RTX 3060 11.845 GB
Processing Settings
device: 0 vram: 1 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis 1X: 10.08 fps 2X: 07.34 fps 4X: 01.90 fps
Iris 1X: 10.19 fps 2X: 05.86 fps 4X: 01.92 fps
Proteus 1X: 10.22 fps 2X: 06.53 fps 4X: 02.00 fps
Gaia 1X: 03.27 fps 2X: 02.23 fps 4X: 01.51 fps
Nyx 1X: 03.84 fps 2X: 03.22 fps
4X Slowmo Apollo: 13.21 fps APFast: 34.18 fps Chronos: 07.49 fps CHFast: 12.66 fps
Topaz Video AI Beta v4.0.1.0.b
System Information
OS: Windows v11.22
CPU: 12th Gen Intel(R) Core(TM) i7-12700H 63.52 GB
GPU: Intel(R) Arc(TM) A770M Graphics 15.875 GB
GPU: Intel(R) Iris(R) Xe Graphics 0.125 GB
Processing Settings
device: 2 vram: 1 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis 1X: 08.81 fps 2X: 03.90 fps 4X: 01.37 fps
Iris 1X: 07.32 fps 2X: 04.31 fps 4X: 01.45 fps
Proteus 1X: 07.66 fps 2X: 03.22 fps 4X: 01.10 fps
Gaia 1X: 03.21 fps 2X: 01.78 fps 4X: 00.28 fps
Nyx 1X: 03.13 fps 2X: 02.73 fps
4X Slowmo Apollo: 06.10 fps APFast: 17.43 fps Chronos: 06.34 fps CHFast: 08.57 fps
1 Like
I have a unique setup because I have two of the same GPU in two machines; one a state-of-the-art AMD workstation, and the other a 14nm Intel, and the results are rather different esp. with the â4Xâ models.
The newer AMD platform on Win11:
Topaz Video AI v4.0.2
System Information
OS: Windows v11.22
CPU: AMD Ryzen 9 7950X 16-Core Processor 63.182 GB
GPU: NVIDIA GeForce RTX 3090 23.77 GB
GPU: AMD Radeon(TM) Graphics 0.47446 GB
Processing Settings
device: 0 vram: 1 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis 1X: 23.63 fps 2X: 12.82 fps 4X: 03.64 fps
Iris 1X: 22.27 fps 2X: 12.96 fps 4X: 04.56 fps
Proteus 1X: 23.29 fps 2X: 11.81 fps 4X: 03.69 fps
Gaia 1X: 08.01 fps 2X: 05.50 fps 4X: 03.63 fps
Nyx 1X: 09.78 fps 2X: 08.15 fps
4X Slowmo Apollo: 31.00 fps APFast: 63.00 fps Chronos: 17.94 fps CHFast: 28.73 fps
The older Intel platform on Win10:
Topaz Video AI v4.0.2
System Information
OS: Windows v10.22
CPU: Intel(R) Core(TM) i7-10700K CPU @ 3.80GHz 31.867 GB
GPU: NVIDIA GeForce RTX 3090 23.77 GB
GPU: NVIDIA GeForce RTX 3090 23.77 GB
GPU: NVIDIA GeForce RTX 3090 23.77 GB
Processing Settings
device: 0 vram: 1 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis 1X: 23.97 fps 2X: 10.63 fps 4X: 02.85 fps
Iris 1X: 22.19 fps 2X: 11.08 fps 4X: 03.00 fps
Proteus 1X: 20.00 fps 2X: 11.88 fps 4X: 02.72 fps
Gaia 1X: 07.94 fps 2X: 05.30 fps 4X: 02.64 fps
Nyx 1X: 09.54 fps 2X: 07.77 fps
4X Slowmo Apollo: 25.48 fps APFast: 46.01 fps Chronos: 17.41 fps CHFast: 23.21 fps
1 Like
Is it possible to test the 7950X with two 3090s?
I wonder whatâs the performance like with the new M3 Max.
1 Like
I realized Iâm on 3.2.9 but see prelim benchmarks on my 1 hour old m3 max 14" MacBook Pro. Not sure how much of. difference is there between the newer update.
Topaz Video AI v3.2.9
System Information
OS: Mac v14.01
CPU: Apple M3 Max 64 GB
GPU: Apple M3 Max 48 GB
Processing Settings
device: 0 vram: 1 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis 1X: 11.72 fps 2X: 07.43 fps 4X: 02.56 fps
Proteus 1X: 12.25 fps 2X: 06.84 fps 4X: 02.19 fps
Gaia 1X: 03.41 fps 2X: 02.37 fps 4X: 01.68 fps
4X Slowmo Apollo: 09.26 fps APFast: 37.45 fps Chronos: 04.02 fps CHFast: 06.64 fps
1 Like
What is the power consumption of M3 Max at maximum Topaz load?
So thatâs about 30% slower than a M2 Ultra 24/60c.
But then Topaz doesnât really scale well with many GPU cores as the M2 Ultra with 60 GPU cores is only about double the speed for many cases as the old M1 Pro with 16 GPU cores.
Topaz Video AI v4.0.3
System Information
OS: Mac v14.01
CPU: Apple M2 Ultra 64 GB
GPU: Apple M2 Ultra 48 GB
Processing Settings
device: 0 vram: 1 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis 1X: 14.98 fps 2X: 08.73 fps 4X: 02.66 fps
Iris 1X: 07.58 fps 2X: 02.74 fps 4X: 02.50 fps
Proteus 1X: 14.78 fps 2X: 07.66 fps 4X: 02.37 fps
Gaia 1X: 05.52 fps 2X: 03.74 fps 4X: 02.69 fps
Nyx 1X: 05.24 fps 2X: 06.44 fps
4X Slowmo Apollo: 18.36 fps APFast: 68.84 fps Chronos: 05.73 fps CHFast: 07.64 fps
NB: the issue with Iris being abysmally slow on Apple Silicon especially with the 2x model is till there - heck on the M1 Pro Iris 4x is even faster than 2x!
Hereâs the 10 GPU core version of the base M3, 24Gb iMac
Topaz Video AI v4.0.3
System Information
OS: Mac v14.0101
CPU: Apple M3 24 GB
GPU: Apple M3 16 GB
Processing Settings
device: 0 vram: 1 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis 1X: 06.20 fps 2X: 04.01 fps 4X: 01.54 fps
Iris 1X: 04.47 fps 2X: 00.62 fps 4X: 00.80 fps
Proteus 1X: 05.64 fps 2X: 03.64 fps 4X: 01.19 fps
Gaia 1X: 02.03 fps 2X: 01.45 fps 4X: 01.05 fps
Nyx 1X: 01.58 fps 2X: 00.95 fps
4X Slowmo Apollo: 08.06 fps APFast: 32.88 fps Chronos: 02.26 fps CHFast: 03.52 fps
2 Likes
Quite nice for the base config.
But Iris 2x⊠ouch! 
Yep there is something about Iris x2 that doesnât agree with Mac Silicon
