Accurate and reliable cancer categorization is crucial for informing medical decisions and improving patient care. Deep learning algorithms have emerged as a promising approach due to their ability to extract intricate patterns and correlations from large clinical datasets. In this paper, we propose a novel ensemble technique based on the Levy Stable probability density function (PDF) and deep learning methodologies for cancer classification, aiming to enhance the accuracy of cancer subtype prediction. Levy Stable model employs a robust ranking mechanism and optimizes for real-time hardware inferencing. This approach is evaluated on the Ham10k dataset. The experimental results demonstrate that the proposed model outperforms state-of-the-art methods in cancer classification. To optimize the model for real-time hardware inferencing, we utilize the OpenVINO toolkit, achieving high-performance inference rates of 110FPS and 91FPS for float16 and float32 precision, respectively, on an Intel i7 CPU.