You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This super task is about writing the cpu renderer for Alpenite. We should first implement DAG-node selection and draw call generation on CPU and move it to GPU only once that works. Some parts probably should stay on CPU anyways (to have a simpler architecture), e.g. generating network requests.
Roughly, we need:
DAG loading over the network. Given one root node, all child nodes should be found and loaded on demand. The root node should contain a good few LOD levels of child nodes. Every child node should again contain additional parts of the DAG (several levels below itself). For eviction, we need a similar strategy as in the 2.5D renderer (no missing parts between the current node and the root). this is mostly nucleus work.
meshlet rendering: write a new geometry rendering pipeline with the gltf meshlets (maybe reuse the prototype renderer?) this is mostly engine work (webgpu or webgl). for this task, we can use a fixed set of meshlets.
scheduled meshlet rendering: create a draw list using the parallel DAG cut algorithm
The text was updated successfully, but these errors were encountered:
This super task is about writing the cpu renderer for Alpenite. We should first implement DAG-node selection and draw call generation on CPU and move it to GPU only once that works. Some parts probably should stay on CPU anyways (to have a simpler architecture), e.g. generating network requests.
Roughly, we need:
The text was updated successfully, but these errors were encountered: