The paper introduces 4D Rotor Gaussian Splatting (4DRotorGS), a novel method for synthesizing novel views of dynamic scenes. Inspired by 3D Gaussian Splatting (3DGS), 4DRotorGS represents dynamic scenes using anisotropic 4D XYZT Gaussians, which are temporally sliced to form 3D Gaussians at each timestamp. This approach effectively captures complex dynamics and fine details, especially for scenes with abrupt motions. The method is implemented in a highly optimized CUDA framework, achieving real-time inference speeds of up to 277 FPS on an RTX 3090 GPU and 583 FPS on an RTX 4090 GPU. Extensive evaluations on diverse datasets demonstrate the superior efficiency and effectiveness of 4DRotorGS, outperforming existing methods in both quantitative and qualitative metrics.The paper introduces 4D Rotor Gaussian Splatting (4DRotorGS), a novel method for synthesizing novel views of dynamic scenes. Inspired by 3D Gaussian Splatting (3DGS), 4DRotorGS represents dynamic scenes using anisotropic 4D XYZT Gaussians, which are temporally sliced to form 3D Gaussians at each timestamp. This approach effectively captures complex dynamics and fine details, especially for scenes with abrupt motions. The method is implemented in a highly optimized CUDA framework, achieving real-time inference speeds of up to 277 FPS on an RTX 3090 GPU and 583 FPS on an RTX 4090 GPU. Extensive evaluations on diverse datasets demonstrate the superior efficiency and effectiveness of 4DRotorGS, outperforming existing methods in both quantitative and qualitative metrics.