i’m interested in that too. i just bought a 3080 yesterday and posted about it on reddit and everyone was asking why I didn’t just buy a 4070 instead. My answer was, because there aren’t any benchmarks on here yet, and also after tax the 4070 comes out to being $100 more then what i paid for the 3080.
I think I will get the RTX 4070 
I am planning to get the Asus Dual White Version. 
It is not available yet, but Asus told me it will arrive at the end of this month.
https://dp-forum.com/t/asus-dual-geforce-rtx-4070-white-12gb-gddr6x/419?u=lhkjacky
I am using a PHANTEKS Eclipse P500A D-RGB Matte White case.
It normally fit 435mm display card, however I have add 8 harddisk trays into the case. Now, it only fit display card less than 300mm. 
Your’re gonna pay a premium for a white color ($680 at least). At that point just spend a few bucks more and get an ASUS TUF 4070 TI for 799.
Sadly, my local computer store is selling at $880 
Also I need to remove 1~2 hhd trays to fit the card, which I try not to.
Anyone got any benchmarks for Laptop version of Nvidia cards. This is the 3070 Laptop:
Topaz Video AI v3.2.1
System Information
OS: Windows v10.2009
CPU: AMD Ryzen 7 5800H with Radeon Graphics 15.356 GB
GPU: NVIDIA GeForce RTX 3070 Laptop GPU 7.8447 GB
GPU: AMD Radeon™ Graphics 0.48492 GB
Processing Settings: device: 0 vram: 1 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis 1X: 9.78 fps 2X: 7.93 fps 4X: 1.98 fps
Proteus 1X: 10.51 fps 2X: 7.26 fps 4X: 1.82 fps
Gaia 1X: 3.25 fps 2X: 2.20 fps 4X: 1.59 fps
4X Slowmo Apollo: 13.84 fps Chronos: 7.81 fps Chronos Fast: 12.93 fps
1 Like
I asked someone in the MacRumors forum to post his benchmarks from his Macs. This is thanks to bnumerick.
Topaz Video AI v3.2.2
System Information
OS: Mac v13.0301
CPU: Apple M1 Ultra 64 GB
GPU: Apple M1 Ultra 48 GB
Processing Settings: device: 0 vram: 1 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis 1X: 13.02 fps 2X: 7.74 fps 4X: 2.87 fps
Proteus 1X: 12.07 fps 2X: 6.86 fps 4X: 2.37 fps
Gaia 1X: 4.87 fps 2X: 3.16 fps 4X: 2.34 fps
4X Slowmo Apollo: 8.48 fps Chronos: 4.12 fps Chronos Fast: 6.11 fps
Here’s my 16" M2 Max (14" is quite a bit slower due to the lower TDP on the system)
Topaz Video AI v3.2.2
System Information
OS: Mac v13.0301
CPU: Apple M2 Max 32 GB
GPU: Apple M2 Max 21.333 GB
Processing Settings: device: 0 vram: 1 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis 1X: 11.28 fps 2X: 6.78 fps 4X: 2.41 fps
Proteus 1X: 10.40 fps 2X: 6.43 fps 4X: 2.15 fps
Gaia 1X: 3.13 fps 2X: 2.27 fps 4X: 1.82 fps
4X Slowmo Apollo: 9.48 fps Chronos: 3.11 fps Chronos Fast: 5.30 fps
Here’s the 16" M1 Max 32 core
Topaz Video AI v3.2.2
System Information
OS: Mac v13.0301
CPU: Apple M1 Max 32 GB
GPU: Apple M1 Max 21.333 GB
Processing Settings: device: 0 vram: 1 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis 1X: 8.43 fps 2X: 5.30 fps 4X: 2.16 fps
Proteus 1X: 9.24 fps 2X: 5.61 fps 4X: 1.91 fps
Gaia 1X: 2.70 fps 2X: 1.66 fps 4X: 1.36 fps
4X Slowmo Apollo: 7.01 fps Chronos: 2.66 fps Chronos Fast: 4.69 fps
I’ve updated the VEAI community performance overview spreadsheet with the results from the previous couple of posts back to @prescott… if anyone wants to put theirs in this will give us a sort-able overview of where things are at…
4 Likes
I bought a MacBook Pro with M2 Max to try it out. Here’s the benchmark.
Topaz Video AI v3.2.2
System Information
OS: Mac v13.0301
CPU: Apple M2 Max 32 GB
GPU: Apple M2 Max 21.333 GB
Processing Settings: device: 0 vram: 1 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis 1X: 11.15 fps 2X: 7.54 fps 4X: 2.95 fps
Proteus 1X: 11.04 fps 2X: 6.94 fps 4X: 2.50 fps
Gaia 1X: 3.40 fps 2X: 2.53 fps 4X: 1.99 fps
4X Slowmo Apollo: 9.73 fps Chronos: 3.63 fps Chronos Fast: 6.09 fps
Topaz Video AI v3.2.2
System Information
OS: Windows v11.2009
CPU: 13th Gen Intel(R) Core™ i9-13900K 63.715 GB
GPU: NVIDIA GeForce RTX 4090 23.59 GB
Processing Settings: device: 0 vram: 1 instances: 0
Input Resolution: 3840x2160
Benchmark Results
Artemis 1X: 8.92 fps 2X: 5.19 fps 4X: 1.52 fps
Proteus 1X: 7.72 fps 2X: 4.78 fps 4X: 1.47 fps
Gaia 1X: 3.33 fps 2X: 2.26 fps 4X: 1.31 fps
4X Slowmo Apollo: 9.92 fps Chronos: 6.86 fps Chronos Fast: 12.97 fps
Topaz Video AI v3.2.2
System Information
OS: Windows v11.2009
CPU: AMD Ryzen 9 3900X 12-Core Processor 47.932 GB
GPU: NVIDIA GeForce RTX 3060 11.845 GB
Processing Settings: device: 0 vram: 1 instances: 0
Input Resolution: 1920x1080
Benchmark Results
Artemis 1X: 9.38 fps 2X: 6.79 fps 4X: 2.20 fps
Proteus 1X: 9.07 fps 2X: 6.54 fps 4X: 2.17 fps
Gaia 1X: 3.11 fps 2X: 2.12 fps 4X: 1.44 fps
4X Slowmo Apollo: 12.48 fps Chronos: 7.42 fps Chronos Fast: 11.98 fps
Is anyone else finding that the benchmarks don’t stack up with actual real world usage? For example. i just got a RTX 3080 today. I upgraded from a RTX 2070 Super. The 1080p 2x Gaia benchmark for my 2070 Super shows 2.54fps, but right now I’m processing a video thats 1080p with 2x Gaia and my fps is 2.8fps. So did I really only gain not even 1fps in processing power by upgrading from a 2070 Super to a 3080? That seems VERY ridiculous. I basically just wasted $560 for less then a half percent increase in power.
But Gaia is slow on all GPUs because, I understand, it relies more on the CPU than the other models so a better GPU won’t improve things much.
Try it on Artemis and Proteus.
If I understand correctly, the 2070s 2.54fps are from benchmark, while the 3080 2.8fps are real-world result.
In my opinion, we can’t directly compare benchmark vs real-world result, because the benchmark does not include encode/decode and data read/write. So the result can vary.
It better to compare 2070s benchmark vs 3080 benchmark,
or 2070s real-world result vs 3080 real-world result.
I’m impressed!
Topaz Video AI v3.2.2
System Information
OS: Windows v11.2009
CPU: 13th Gen Intel(R) Core(TM) i9-13900K 31.787 GB
GPU: NVIDIA GeForce RTX 4070 Ti 11.729 GB
GPU: Intel(R) UHD Graphics 0.125 GB
Processing Settings: device: 0 vram: 1 instances: 0
Input Resolution: 1920x1080
Benchmark Results
Artemis 1X: 24.23 fps 2X: 16.56 fps 4X: 5.03 fps
Proteus 1X: 22.23 fps 2X: 15.07 fps 4X: 5.06 fps
Gaia 1X: 8.23 fps 2X: 5.74 fps 4X: 3.78 fps
4X Slowmo Apollo: 35.92 fps Chronos: 18.56 fps Chronos Fast: 28.25 fps
Other way around. Gaia relies on something more like rasterization power of the GPU.
Well I don’t understand that lol but OK, I thought that was what I’d read on here. I just avoid Gaia because it’s so SLOOWWWW on my 3060 ti compared to Artemis and Proteus.
I avoid it because I don’t usually see much difference in it and a simple Lanczos filter upscale.
1 Like
Topaz Video AI v3.2.3
System Information
OS: Windows v11.2009
CPU: AMD Ryzen 9 7950X 16-Core Processor 63.14 GB
GPU: NVIDIA GeForce RTX 4090 23.59 GB
GPU: AMD Radeon(TM) Graphics 0.47446 GB
Processing Settings: device: 0 vram: 1 instances: 1
Input Resolution: 1920x1080
Benchmark Results
Artemis 1X: 42.74 fps 2X: 15.43 fps 4X: 4.04 fps
Proteus 1X: 36.84 fps 2X: 18.44 fps 4X: 4.93 fps
Gaia 1X: 15.68 fps 2X: 10.47 fps 4X: 5.16 fps
4X Slowmo Apollo: 42.81 fps APFast: 63.90 fps Chronos: 32.07 fps CHFast: 36.38 fps
Topaz Video AI v3.2.2
System Information
OS: Mac v13.0301
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: 10.98 fps 2X: 7.27 fps 4X: 2.70 fps
Proteus 1X: 11.17 fps 2X: 7.17 fps 4X: 2.37 fps
Gaia 1X: 3.37 fps 2X: 2.53 fps 4X: 2.04 fps
4X Slowmo Apollo: 9.71 fps Chronos: 3.69 fps Chronos Fast: 6.01 fps
Topaz Video AI v3.2.3
System Information
OS: Mac v13.0301
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: 10.94 fps 2X: 7.52 fps 4X: 2.64 fps
Proteus 1X: 10.81 fps 2X: 7.11 fps 4X: 2.50 fps
Gaia 1X: 3.38 fps 2X: 2.52 fps 4X: 1.99 fps
4X Slowmo Apollo: 9.15 fps APFast: 31.15 fps Chronos: 3.65 fps CHFast: 6.01 fps
