Would you be interested in GPU acceleration and DirectCompute tutorial?

So im wondering if a payed video tutiorial with examples would sell around here and really how much could it fetch?

 

You would likely need free tutorials to tempt people to pay for the rest. If the free ones were good you might get some people interested in paying for the rest. Do note that this is a relative small market:

People who want GPU acceleration, who can't do it themselves already or work out how from free resources, who are willing to pay for tutorials, and who want video tutorials.

 

Additionally they would need to be people who don't mind being restricted to Windows.

 

Yeah, I'm interested and I would pay. Maybe in the $20 range (or more if it was really good quality).

 

You lost me at video. This does not sound like a topic for video... unless this is leaning toward theory, at which point you lost me at paid.

 

I would probably appreciate skeleton project with blank GPU compute "shader", if it is free. Because something like that could save me 30 minutes of research. Heck, I probably wouldn't need any explanations, just skeleton code.

I would not pay for a tutorial. I would not watch a video tutorial either.

I dislike video tutorials, because you can't Ctrl+F through them, can't scan the information for relevant info, and people who make tutorials sometiimes ramble about unrelated topics, wasting my time.

Nothing personal, of course. You asked for opinion, so here's one.

 

There is a complete collection of unity tutorials for various types of games, it's $20 and it has like a 100 videos. (I can't find the link now)
You're aiming at a way smaller market with probably just a few videos if you're going to explain just the basics of how it works.
And you're both competing against free alternatives.

Now what could make this tempting is if you would add tutorials on solving practical problems that could be solved much faster by using mass-parallelism like: cracking passwords (trying many combinations at the same time), calculating outcome for neural networks (they can have hundreds of nodes who's value is calculated independently) or physics that involves a lot of colliders, stuff like that.