Constructive feedback: Utilize more of my hadware for processing.
My settings: Max vram and I selected my GPU as the processor
Specs:
My CPU is an i7 - 7820HK
GPU is a GTX 1070 8GB vram
32GB Ram
1TB NVME
Processing a video with my GPU is at an average range of about 4%, peak at 18%, and low at 2%
GPU VRam used is 3.4 of 8
My cpu bounced around between 8% - 45%
My ram is being utilized within a range of 2.5GB and 3.5GB.
My suggestion is to have my gpu more utilized like 90%-100% to have my videos process faster since its only using a small amount of my gpus processing power and if ram could help speed up the process I have considerably more ram and my cpu could handle a bit more as well.
I found that task manager doesn’t reflect the true GPU usage as it does with the Nvidia app. My Asus desktop has its Armory app linked to the Nvidia GPU and it is showing 80-90% usage of my GTX-1650, whereas the task manager only shows like 5-10%. I’ve included a snapshot of my Asus armory app showing a more realistic representation of my CPU and GPU usage. VEAIpicture hosted at ImgBB — ImgBB
Thanks for letting me know, I tried the nvidia app and the xbox game bar app and noticed they both show my gpu and cpu are used more than I thought they were. I am actually seeing it peak to 100%. My vram is at 40% and my ram is 29%.
With my GTX-1650, the difference between 1.5.0 and 2.6.0 is very noticeable. I beta tested VEAI awhile back, but gave up @ 1.5.0 because it was too slow and a bit buggy.
Tried VEAI again this year recently and now bought the 2.6.0 version because it was delivering way faster performance and seemed to utilize my GPU well finally.
In 1.5.0, 480p-720p was like 0.88sec/frame. In 2.6.0, 480p-1080p is like 0.35sec/frame
In 1.5.0 480p-2160p was like 3sec/frame. In 2.6.0, 480p-2160p is like 0.88sec/frame
So something is cooking, because it used to take days with 1.5.0, but now take hours with the current conversion speed, which is more to my liking with the GTX-1650.
VRAM usage is roughly 50-70% depending on upscale size and model. RAM usage is similar to yours.
