Do you know the max on a 12Gig gpu?
If not then I would appreciate datasets for testing, over 20mp. JPG’s!
Do you know the max on a 12Gig gpu?
If not then I would appreciate datasets for testing, over 20mp. JPG’s!
Have you taken a look through our ODMData repository?
Thanks! I’ve forgotten about that.
Cant find anything good there. Dont you have a Fuji 26mp?
I increased image size to 29mp and processed, it used my 12 gpu without problems.
So I’m going for a Fuji TX-3 next!
24MP, yeah!
Trying to get time to go out and do a ground-level collection. Found a really neat piece of construction equipment sitting out.
I’m going to use my Olympus for the time being, there’s a more suitable lens on the way, the Nisi 9mm.
A dream would be the XT-3 with a zeiss touit 12/2.8 So now I need to sell some stuff and save some cash to make it real.
That screenshot is during point cloud densification when processing 20MP images. The yellow bars on the left scale show VRAM usage, the red line on the right scale shows GPU usage.
When using 12MP with GPU point cloud densification ~450Mbyte of VRAM were used during that stage, when computing with 20MP and GPU point cloud densification up to 900Mbyte were used.
Does not look like you should run into any limitations with 12Gbyte of VRAM there
it’s been said that you need 8gig vram for 20mp images
Point cloud densification or feature detection?
During point cloud rendering the GPU is used to 100% but the VRAM stays super low (<1Gbyte) for all processes that I have run.
12Gig vram isn’t enuf for future detection and 6 gig wasn’t enuf for pointcloud densification
Interesting.
I use a Geforce 1660 with 6Gbyte of VRAM and it can do point cloud densification, often on ultra and the VRAM usage is less than 1Gbyte.
But feature-detection has some funny VRAM limitation.
That’s strange, what Megapixel do you have precisely?
12MP (4000x3000) from a Phantom 3 Pro and 20MP (5472x3648) from a Phantom 4 Pro.