I mean the real power consumption of the whole system measured with a device at the socket. Not what some fancy software wants to tell you.
No way a PC will only roughly compete with an Apple Silicon rig here.
From Apples Support site.
Mac Studio (2023)
M2 Ultra 24-core CPU and 76-core GPU, 192GB shared memory, 8TB SSD
Power consumption
Idle / Max
10 W / 295 W
So half the power half the speed.
As said: Take a measuring device and plug it directly into the outlet on the wall, then plug in the computer. The Mac studio (with 60 GPU cores here) idles at under 10W and also uses 10W in a typical desktop use scenario: Word, Excel, Webbrowsing, RDP.
Iâve seen a maximum of about 220W, yes, but not with TVAI alone, but only when I had two simultaneous TVAI encodings and at the same time three handbrake jobs, one CPU, two the specialized h265 hw to max out this component as well. TVAI alone is roughly 150-160W.
I have lots of PCs of varying performance, notebooks and Macs, both intel and AppleSilicon here and did several measurements in idle / full throttle scenarios. The Apple silicon machines are lower in power consumption than âclassic PCâ notebook technology.
Which you can already easily see by the small form factor of a Mac mini/studio - no way getting the thermal footprint of a halfway performing desktop PC into such a small device
And that (measuring directly at the outlet) is the only measurement that you can rely on.
Everything other is as if youâd just swallow those fuel consumption values the car manufacturers tell you without measuring it yourself (or even the estimated range of electric cars 
)
But then this discussion is going really off-topic nowâŠ
Donât forget to plug in the NAS (with a CPU, RAM and Cache SSD, too when measuring, bc iâve 15 Terabyte inside the case and dont need a NAS.
There are some people here in the forum who run their PCs in power saving mode and donât have a big drop in performance, I also use my 7950X in 125 watt mode and donât notice anything about it being slower, itâs even 63% faster on certain tasks than my big 24 core 3960X with four channel RAM with 128 GB.
Oh yes.
If the job takes 8 hours, youâre done after 4 hours with the fast system.
Not to mention that with the 30XX and 40XX series you can also do other things on the side and the device does not become a single use tool.
1 Like
Time is money. Take also overall time cost in calculation - not only summary power consumption.
Topaz Video AI v3.4.0
System Information
OS: Mac v13.05
CPU: Apple M2 Max 96 GB
GPU: Apple M2 Max 72 GB
Processing Settings
device: 0 vram: 1 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis 1X: 11.23 fps 2X: 07.33 fps 4X: 02.53 fps
Iris 1X: 11.54 fps 2X: 01.67 fps 4X: 01.03 fps
Proteus 1X: 11.19 fps 2X: 07.29 fps 4X: 02.46 fps
Gaia 1X: 03.43 fps 2X: 02.53 fps 4X: 02.04 fps
4X Slowmo Apollo: 12.00 fps APFast: 51.84 fps Chronos: 03.70 fps CHFast: 06.18 fps
Topaz Video AI v3.4.0
System Information
OS: Windows v10.22
CPU: AMD Ryzen 7 3700X 8-Core Processor 63.952 GB
GPU: NVIDIA GeForce RTX 4090 23.59 GB
Processing Settings
device: 0 vram: 1 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis 1X: 32.21 fps 2X: 09.36 fps 4X: 02.48 fps
Iris 1X: 27.71 fps 2X: 09.48 fps 4X: 02.48 fps
Proteus 1X: 22.26 fps 2X: 08.90 fps 4X: 02.42 fps
Gaia 1X: 15.05 fps 2X: 08.76 fps 4X: 02.41 fps
4X Slowmo Apollo: 32.31 fps APFast: 42.62 fps Chronos: 30.01 fps CHFast: 26.04 fps
NVIDIA Studio Driver 536.99
NTSC SD input
============
Topaz Video AI v3.4.0
System Information
OS: Windows v11.22
CPU: 12th Gen Intel(R) Core(TM) i9-12900KF 31.836 GB
GPU: NVIDIA GeForce RTX 3080 Ti 11.816 GB
Processing Settings
device: 0 vram: 1 instances: 0
Input Resolution: 720x480
Benchmark Results
Artemis 1X: 131.03 fps 2X: 76.95 fps 4X: 24.59 fps
Iris 1X: 96.54 fps 2X: 52.13 fps 4X: 20.72 fps
Proteus 1X: 120.03 fps 2X: 70.09 fps 4X: 23.03 fps
Gaia 1X: 46.99 fps 2X: 33.77 fps 4X: 24.08 fps
4X Slowmo Apollo: 128.20 fps APFast: 290.55 fps Chronos: 103.38 fps CHFast: 128.34 fps
FHD Input
=========
Topaz Video AI v3.4.0
System Information
OS: Windows v11.22
CPU: 12th Gen Intel(R) Core(TM) i9-12900KF 31.836 GB
GPU: NVIDIA GeForce RTX 3080 Ti 11.816 GB
Processing Settings
device: 0 vram: 1 instances: 0
Input Resolution: 1920x1080
Benchmark Results
Artemis 1X: 24.07 fps 2X: 17.15 fps 4X: 04.42 fps
Iris 1X: 17.85 fps 2X: 09.86 fps 4X: 03.43 fps
Proteus 1X: 22.63 fps 2X: 15.75 fps 4X: 04.61 fps
Gaia 1X: 08.49 fps 2X: 05.85 fps 4X: 03.97 fps
4X Slowmo Apollo: 33.47 fps APFast: 70.82 fps Chronos: 19.19 fps CHFast: 29.04 fps
Topaz Video AI v3.3.4
System Information
OS: Windows v11.22
CPU: Intel(R) Xeon(R) CPU E5-2690 v4 @ 2.60GHz 63.909 GB "14 cores running @3.2 GHz during test
GPU: NVIDIA GeForce RTX 2080 Ti 10.782 GB
Processing Settings
device: 0 vram: 1 instances: 0
Input Resolution: 1920x1080
Benchmark Results
Artemis 1X: 15.01 fps 2X: 08.52 fps 4X: 02.40 fps
Proteus 1X: 14.70 fps 2X: 09.61 fps 4X: 02.36 fps
Gaia 1X: 04.77 fps 2X: 03.20 fps 4X: 02.23 fps
4X Slowmo Apollo: 06.31 fps APFast: 36.96 fps Chronos: 10.51 fps CHFast: 12.49 fps
Is there a summary of all benchmarks in place, there I can see the results for
a) the different Intel and Ryzen processors?
b) the different Nvidia cards?
c) combination of Intel processor and Nivida cards?
d) rendert Input Resolution?
Or could I filter the results?
The benchmarks are very ecxiting to make a buy/invest decision.
2 Likes
Results after adjusting motherboard power delivery to reduce crashes. Incredibly, there seems to be this sweet spot where saving energy also improves performance.
Topaz Video AI v3.3.2
System Information
OS: Windows v11.22
CPU: AMD Ryzen 7 5800X 8-Core Processor 15.929 GB
GPU: NVIDIA GeForce RTX 3070 Ti 7.8428 GB
Processing Settings
device: 0 vram: 0.9 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis 1X: 15.71 fps 2X: 08.65 fps 4X: 02.22 fps
Proteus 1X: 14.36 fps 2X: 08.25 fps 4X: 02.20 fps
Gaia 1X: 05.24 fps 2X: 03.60 fps 4X: 02.34 fps
4X Slowmo Apollo: 19.45 fps APFast: 35.86 fps Chronos: 12.40 fps CHFast: 19.07 fps
Apple iMac Pro 2017 max spec 18 core CPU and almost max spec GPU
Topaz Video AI v3.4.3
System Information
OS: Mac v13.0501
CPU: Intel(R) Xeon(R) W-2191B CPU @ 2.30GHz 128 GB
GPU: AMD Radeon Pro Vega 64 15.984 GB
Processing Settings
device: 0 vram: 1 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis 1X: 03.07 fps 2X: 01.78 fps 4X: 00.63 fps
Iris 1X: 02.26 fps 2X: 01.17 fps 4X: 00.38 fps
Proteus 1X: 03.12 fps 2X: 02.04 fps 4X: 00.80 fps
Gaia 1X: 01.12 fps 2X: 00.72 fps 4X: 00.53 fps
4X Slowmo Apollo: 02.10 fps APFast: 08.22 fps Chronos: 00.28 fps CHFast: 00.95 fps
