Point Cloud Mamba (PCM) is a novel method for 3D point cloud analysis that leverages state space models (SSMs), specifically the Mamba architecture, to achieve superior performance compared to existing point-based methods. The method introduces a Consistent Traverse Serialization (CTS) technique to convert 3D point clouds into 1D sequences while preserving spatial adjacency between points. This allows Mamba to effectively model point cloud data with linear computational complexity. PCM also incorporates order prompts to guide Mamba in handling different point sequence orders and positional encoding based on spatial coordinate mapping to inject positional information into the sequences. PCM outperforms the state-of-the-art point-based method PointNeXt on multiple datasets, including ScanObjectNN, ModelNet40, ShapeNetPart, and S3DIS. When using a more powerful local feature extraction module, PCM achieves 82.6 mIoU on the S3DIS dataset, surpassing previous methods by significant margins. The method's architecture combines local and global modeling capabilities, enabling it to capture both local and global features of point clouds effectively. PCM's key contributions include the introduction of CTS, order prompts, and positional encoding to enhance Mamba's ability to process point cloud data. The method demonstrates significant improvements in performance across various tasks, including 3D object classification, part segmentation, and semantic segmentation. The results show that PCM achieves state-of-the-art performance on multiple benchmark datasets, highlighting the effectiveness of using Mamba for point cloud analysis. The method also shows potential for further improvements by combining local feature extractors with Mamba layers to better model both local and global features of point clouds.Point Cloud Mamba (PCM) is a novel method for 3D point cloud analysis that leverages state space models (SSMs), specifically the Mamba architecture, to achieve superior performance compared to existing point-based methods. The method introduces a Consistent Traverse Serialization (CTS) technique to convert 3D point clouds into 1D sequences while preserving spatial adjacency between points. This allows Mamba to effectively model point cloud data with linear computational complexity. PCM also incorporates order prompts to guide Mamba in handling different point sequence orders and positional encoding based on spatial coordinate mapping to inject positional information into the sequences. PCM outperforms the state-of-the-art point-based method PointNeXt on multiple datasets, including ScanObjectNN, ModelNet40, ShapeNetPart, and S3DIS. When using a more powerful local feature extraction module, PCM achieves 82.6 mIoU on the S3DIS dataset, surpassing previous methods by significant margins. The method's architecture combines local and global modeling capabilities, enabling it to capture both local and global features of point clouds effectively. PCM's key contributions include the introduction of CTS, order prompts, and positional encoding to enhance Mamba's ability to process point cloud data. The method demonstrates significant improvements in performance across various tasks, including 3D object classification, part segmentation, and semantic segmentation. The results show that PCM achieves state-of-the-art performance on multiple benchmark datasets, highlighting the effectiveness of using Mamba for point cloud analysis. The method also shows potential for further improvements by combining local feature extractors with Mamba layers to better model both local and global features of point clouds.