NyuziProcessor VS catboost

There's also Nyuzi which is more GPGPU focused https://github.com/jbush001/NyuziProcessor, but the author also experimented with having it do 3D graphics.

You could probably layer a software implementation of the rasterization pipeline on top of a compute-focused open-source GPU architecture like Nyuzi: https://github.com/jbush001/NyuziProcessor/

I would expect a 2x slowdown over hardware rasterization, based on NVIDIA's work on such an approach, but this is probably fine if you're just trying to match Voodoo3 performance. And one could imagine bolting a minimal hardware rasterizer on top of Nyuzi to speed things up once the software implementation is working.

Sounds like the Nyuzi processor :D

CatBoost is another popular and high-performance open-source implementation of the Gradient Boosting Decision Tree (GBDT). To learn how to use this algorithm, please see example notebooks for Classification and Regression.

Here are our benchmarks on training time comparing Tangram's Gradient Boosted Decision Tree Library to LightGBM, XGBoost, CatBoost, and sklearn.

Catboost - CatBoost is an open-source software library developed by Yandex. It provides a gradient boosting framework which attempts to solve for Categorical features using a permutation driven alternative compared to the classical algorithm. Link - https://catboost.ai/

CatBoost is an open source algorithm based on gradient boosted decision trees. It supports numerical, categorical and text features. Check out the docs.

If you considering GBDT check out catboost, unfortunately RF mode is not available but library implement lots of interesting categorical encoding tricks that boost accuracy.

The data contains a large number of categorical features. The most suitable for obtaining a base-line model, in my opinion, is CatBoost. It is a high-performance, open-source library for gradient boosting on decision trees.