Video
CPUs, GPUs, TPUs, NPUs, and VPUs Explained – What Do They Do?
A guide to CPUs, GPUs, TPUs, NPUs, and VPUs, explaining how specialised processors divide work across general computing, graphics, machine learning, and vision tasks.
Video summary
This guide explains why different processors exist within modern computing and AI systems. CPUs are presented as general-purpose workhorses, while specialised chips are designed for repeated, high-speed calculations in areas such as graphics, machine learning and computer vision.
The video contrasts GPUs, which render large numbers of graphical calculations in parallel, with neural processing units and other application-specific hardware. It introduces TPUs and VPUs as further examples of chips shaped around particular workloads, helping viewers connect processor names to the kinds of tasks they are intended to accelerate.