HPC (High Performance Compute)
A practice of aggregating computing power in a way that delivers significantly higher performance than an average desktop or workstation.
HPC (High Performance Computing), as the name suggests, is concerned with running advanced application programs efficiently and rapidly. In the field of AI, HPC can provide the necessary computing power to handle large amounts of data, run complex machine learning algorithms, or support simultaneous tasks. High-performance computing clusters are often used in AI research and application for their capability to process and analyze vast data sets at a high speed, enabling faster and more accurate outcomes. For instance, in deep learning, one area of AI, HPC can expedite the training process of neural networks.
The practice of high-performance computing started as early as the 1960s with supercomputers. The term 'High Performance Computing' was coined in the 1990s, and its significance grew along with advances in computing technology and AI.
Many scientists and engineers have contributed to the evolution of HPC systems. Companies like IBM and Cray, alongside organizations like the National Center for Supercomputing Applications (NCSA) and the Department of Energy (DOE) in the United States, have been key players in the development of high-performance computing.