The demand for efficient and high-performance computing systems has led to the development of photonic-based technologies for machine learning. One of the key components of these systems is the photonic tensor core, which performs matrix operations at high speed and low power consumption. In this article, we review the features of photonic tensor cores and their construction for use in neural networks. We discuss the advantages of photonic-based technologies over traditional electronic-based systems, as well as the challenges in their implementation. We also highlight recent advancements in the development of photonic tensor cores for machine learning applications.