I meant if it crashes or something later on. Very few users who run on CPU or integrated graphics, so if you have any issues now or in the future let us know.
I will certainly do
1 Like
woohoo! Thanks guys.
So after a full DDU driver wipe things are now running as expect, though i did notice my 2x 2080ti with NVLINK do not show up as seperate GPUās even in the 1.8 BETA. But this is likely just due to Nvlink not playing nice with VEAI, it only lists a single 2080ti and my 3090. But hey this is a step forward at least after the full driver wipe.
1 Like
Has anyone done a Gaia-CG 1.6.1 to Artemis-HQv9 1.8.0b comparision using clean high quality 720p or 1080p? Iām doing that tomorrow.
Iāve started this test. Upscaling 720p 200% using Artemis-HQv9 1.8.0b at 0.46sec/frame. Gaia-CG 1.6.1 was 0.9sec/frame. Preview quality looks about the same as Gaia-CG. Iāll know more after the upscaling finishes and the mp4 encoding is done.
What is the quality of Artemis-HQv9? Actually, I care more about Artemis-LQv9.Because Artemis-LQ is better than Artemis-HQ in the current version
Well thatās why I asked if anyone did any testing to confirm what you said. Since it seems no one has, Iām doing that test. The first video is done and I used Gaia-CG from 1.6.1 and Iām very happy with it. Iām testing if the new Artemis-HQv9 can do same same or better quality in half the time. I am curious how Artemis-LQv9 and Artemis-MQv9 does with this video, so that may be next.
I have about 1100 movies, gigs, and TV shows to process. And folks might wonder why I built a second VEAI systemā¦ Really I want 30 VEAI systems and lots of automation + several Ryzen 3990X boxes. Iām working on the automation part now. Iām not a fast programmer, so itās at least 3 to 6 months away from working. And I have to get a job as my savings are almost depleted (my company majorly downsized while I was out sick with Covid19, which is fine as I disliked the place, not the work, the placeā¦).
Based on my tests yesterday with Artemis-LQv9 and Artemis-MQv9 in 1.8.0b, I might conclude Artemis-MQv9 gives more pleasing results with noisy video. I need to do more watching and encoding with different sharpening settings to decide.
For my own use and my eye, Artemis LQ V9 is the appropriate replacement for the older Gaia CG. There are plenty of sources that donāt/didnāt look good with either of those, and then MQ and HQ or one of the other models are the best choice. But, if it looked best with CG in 1.6.1 it looks best to me with LQ V9 now.
Hello can anyone tell me what performs better the 6900xt or the 3090?
we need a benchmark for this kinda stuff xd
2 Likes
Use 1.6.1 version, it should be way faster.
My system. The 3090 is as fast as the old 2080, or even slower. I donāt know where the bottleneck is.
i9-10900K
Gigabyte Z490 AORUS MASTER
DDR4-3200 Corsair 32G
U.S. commercial pirate ship RM850x
On my systems:
720P to 2K(200%) Artemis v8
RTX2080:0.22~0.26sec/frame
RTX3090:0.22~0.28sec/frame
1 Like
Thatās kind of off putting thought the new update would show improvement for 3080/3090/6800xt/6900xt, Guess Iāll use other AI until this gets fixed.
Thereās another video AI like VEAI? Whatās the name of the program?
I donāt know if we can mention it here 
but look for the āPainter of Mona Lisaā that will āsolve thingsā
Thanks Topaz for the final v1.8.0. I will be testing today 
Hello, fo you know if thatās for every model? Or only Artemis? (I could only find in the facebook group a reference for Artemis citing this 239% thing)
Itās across the board for all models. You can easily test it this way: Preview a video at 239% upscale and note the seconds per frame. Now, preview the same video at 240% upscale and also take note of the seconds per frame. The latter number will jump considerably higher and will stay at that level at any higher percentage upscale you set. That is because VEAI has switched from 200% enhance mode to 400% enhance mode (which is the highest enhance it is capable of).
VEAI can enhance at 100%, 200% and 400%. All other sizes are resized conventionally using Laczos scaling.
so how huge is the time saving with faster models?
5% 10% 30% ?
