Beginner · Systems
Compute
Visual diagram · (in preparation) · Math · (in preparation) · Worked example · 3 difficulty levels.
TL;DR. The processing power and resources required to train or run machine learning models.
Technical Definition
The processing power and resources required to train or run machine learning models.
How it works
Compute refers to the computational resources, such as processing power (CPUs, GPUs, TPUs), memory, and storage, that are consumed by machine learning models and systems. Training complex models, especially deep neural networks, requires significant compute resources. Inference, the process of using a trained model to make predictions, also demands compute, though typically less than training.
Related Concepts
- GPU — Graphics Processing Unit — a massively parallel processor that powers most modern AI workloads.
- TPU — Tensor Processing Unit — Google's custom-designed AI accelerator chip.
- CPU — Central Processing Unit — the general-purpose chip that runs most software.