To begin, GPU stands for Graphic Processing Unit, and it’s a hardware that was originally developed for the video game world. GPUs were designed to speed up 3D graphics processing, with the capacity to process multiple, parallel computations simultaneously.
Over time, GPUs became more flexible and programmable, leading to their title as the unsung hero of the Artificial Intelligence (AI) revolution.
Graphics rendering and deep learning both deal with an immense number of parallel simultaneous computations per second. The parallel architecture, and the huge memory bandwidth of GPUs make it the perfect solution for compute-intensive deep learning and machine learning AI.
Today, GPUs are an essential part of modern AI infrastructure, and new GPU’s have been created and optimized precisely for AI processes.
This is a great example of how innovations in one field may drastically improve another completely unrelated, yet important technological advancement!
Curious to learn more about GPUs and deep learning? Contact us at info@alturis.ai or use our contact form!