acerola_t
patreon
Project Update and Delay
🕑 Added 2024-09-11 06:20:38 +0000 UTCHello everyone,
I've been hard at work on the particle project again lately now that I'm finished with my Nvidia obligations and have obtained the all powerful 4090. My new pc is built with a dual gpu motherboard so I can benchmark with both the 1660 and the 4090, ensuring my shaders stay grounded for the masses.
So far the project is going really well, but because I am once again traveling this month (for twitch con) I have decided to delay the video to October for multiple reasons. Firstly, the video is going to require quite a lot of linear algebra lecturing as this is all driven by novel matrix multiplications so understanding transformations is necessary and I do not want to cut corners on that educational material. Secondly, I am releasing the channel's first official merch with this video, and so I want the subject matter to be as interesting as possible to demonstrate what I can do when I have more time to work on videos. Lastly, there are a looooot of performance experiments to be done in regards to optimal particle rendering (without the crazy compute splatting pipeline) and I just want the video to be comprehensive and not feel incomplete because as we all know I refuse to make explicit two parters.
I'm sorry for the delay but I promise the wait will be worth it -- I'm really looking forward to putting it all together. Thank you for your support as always!
Comments
Acerola
Yeah particles are tough cause no matter what you're taking the 2x2 pixel overhead so the compute splatting subverts this cost, I don't have time to implement that though so instead I'm trying to find the fastest instancing methods with basic indirect rendering.
helviett
Cool! 'Comprehensive' (and memes) is what people love in your videos in my opinion so it's worth the delay for sure. For my first graphics programming interview I had to make any test project to show my skills. I've decided to go with GPU based Particle System to learn compute shaders on the way. I had tight deadlines so I just picked few GDC talks and codebases\article that share same ideas on PS implementation. I got the job but the interviewer literally said that it's the worst possible GPU based implementation and it's a shame that this implementation is widely adopted :) .
