This study addresses the pressing need for an accurate and efficient automated system for the detection of cancer cells, focusing on Acute Lymphoblastic Leukemia (ALL), the most prevalent form of childhood cancer accounting for 25% of pediatric cases. The dataset comprises images from 118 patients, with meticulous segmentation of cells. These images reflect real-world scenarios, incorporating staining noise and illumination errors, though mitigated during acquisition. Identifying immature leukemic blasts from normal cells is a significant challenge due to morphological similarities. Ground truth labels were expertly annotated by an oncologist, classifying the dataset into two classes: Normal cells and Leukemia blasts. Our methodology involves training a Convolutional Neural Network (CNN) with the multiple architecture to recognize distinctive features associated with leukemia. The intricate task of cell classification is vital for timely and accurate diagnosis, especially given the morphological nuances present in the dataset. Subsequently, the trained model is rigorously tested, evaluating its efficacy in distinguishing between normal and leukemia cells. To optimize testing efficiency, we implement Open MPI, distributing the dataset across multiple processes for parallelized computation. Each process operates concurrently, applying the broadcasted model to expedite the detection process.