PointMamba is a simple state space model for point cloud analysis, inspired by the Mamba model from natural language processing. It uses space-filling curves for point tokenization and a non-hierarchical Mamba encoder to achieve global modeling with linear complexity. The model outperforms Transformer-based methods in terms of performance and efficiency, reducing GPU memory usage and computational costs. PointMamba is designed to be simple and effective, with a focus on achieving superior performance across various point cloud analysis datasets. The model uses a serialization-based mask modeling paradigm, which allows for efficient pre-training and reconstruction of masked point clouds. Experiments show that PointMamba achieves state-of-the-art results on tasks such as object classification, part segmentation, and few-shot learning. The model's linear complexity and efficient design make it a promising alternative to Transformers for 3D vision tasks. PointMamba demonstrates the potential of state space models in point cloud analysis, offering a simple yet effective baseline for future research.PointMamba is a simple state space model for point cloud analysis, inspired by the Mamba model from natural language processing. It uses space-filling curves for point tokenization and a non-hierarchical Mamba encoder to achieve global modeling with linear complexity. The model outperforms Transformer-based methods in terms of performance and efficiency, reducing GPU memory usage and computational costs. PointMamba is designed to be simple and effective, with a focus on achieving superior performance across various point cloud analysis datasets. The model uses a serialization-based mask modeling paradigm, which allows for efficient pre-training and reconstruction of masked point clouds. Experiments show that PointMamba achieves state-of-the-art results on tasks such as object classification, part segmentation, and few-shot learning. The model's linear complexity and efficient design make it a promising alternative to Transformers for 3D vision tasks. PointMamba demonstrates the potential of state space models in point cloud analysis, offering a simple yet effective baseline for future research.