I would have preferred if they had used the die space for the GPU, not this bullshit.

You can use it for blurring your background on your camera

I got an Asus Vivobook with a pretty new AMD chipset that has one... running Fedora, no Idea how to make something use it either :/

I wonder if they have implemented NPU support into Folding@Home yet.

You need a model compiled for the architecture. I saw some for the RK35xx devices when shopping for hardware. I do not think there is software made to split up or run models in general on a NPU. The models must be configured for the physical hardware topology. The stuff that runs on most devices is very small, and these either need a ton of custom fine tuning or they are barely capable of simple tasks.

On the other hand, segmentation models are small, and that makes layers, object identification, and background removal stuff work. Looking at your CPU speed, and available memory, it is unlikely to make much difference. You are also memory constrained for running models, though you could use deepspeed to load from a disk drive too.

The only thing I can think of is AI video game upscalers. Other than that, yeah, it’s a waste of silicon.

https://github.com/FastFlowLM/FastFlowLM

That RTX 4050 is probably even faster though.