Glossary

What is GPU? (Graphics Processing Unit)

A processor that can perform many calculations at the same time, originally developed for graphics but widely used in AI training and inference.

A GPU relies on thousands of small processing units working at the same time. This parallel structure is far better suited than general-purpose processors to the dense matrix operations of machine learning; for this reason it is the core hardware for AI workloads.

In selection, the amount of memory, memory bandwidth, the method of connecting multiple cards to each other and software support are decisive. For large models such as an LLM, memory is often the most critical limit. Training and inference create different hardware requirements.

Buying GPUs brings high investment and operating burden; the rental model provides flexibility. For projects with uncertain usage intensity, GPU as a service may be preferred. This decision matters for teams that want to run an organization-specific model or do fine-tuning.

← All terms

Related solutions

Related articles

Let's plan your critical systems together

Tell us what you need; our team will work out the right architecture and migration path with you.