This paper explores the simulation of fluid flow behavior around an oscillating object in a channel using the Lattice Boltzmann Method (LBM) on both CPU and GPU devices. The study aims to evaluate the effectiveness of the LBM in simulating such flows, particularly focusing on the impact of different Reynolds numbers and oscillation parameters. The authors implemented the LBM using the TensorFlow framework, which allows for efficient parallel processing on GPU devices. The simulation results for a circular cylinder with a radius of 0.00625 and different oscillation frequencies (0.05 Hz, 0.1 Hz, and 0.2 Hz) were compared to known experimental data. The findings show that oscillations disrupt the symmetry of the vortices, leading to larger vortex zones and increased vortex frequency. Performance analysis on a CPU and GPU setup demonstrated significant speedup when using GPU acceleration, achieving a double increase in efficiency. The study concludes that the LBM is well-suited for simulating fluid dynamics around oscillating objects, and GPU acceleration significantly enhances computational efficiency.This paper explores the simulation of fluid flow behavior around an oscillating object in a channel using the Lattice Boltzmann Method (LBM) on both CPU and GPU devices. The study aims to evaluate the effectiveness of the LBM in simulating such flows, particularly focusing on the impact of different Reynolds numbers and oscillation parameters. The authors implemented the LBM using the TensorFlow framework, which allows for efficient parallel processing on GPU devices. The simulation results for a circular cylinder with a radius of 0.00625 and different oscillation frequencies (0.05 Hz, 0.1 Hz, and 0.2 Hz) were compared to known experimental data. The findings show that oscillations disrupt the symmetry of the vortices, leading to larger vortex zones and increased vortex frequency. Performance analysis on a CPU and GPU setup demonstrated significant speedup when using GPU acceleration, achieving a double increase in efficiency. The study concludes that the LBM is well-suited for simulating fluid dynamics around oscillating objects, and GPU acceleration significantly enhances computational efficiency.