What is GPU?
GPU (Graphics Processing Unit): an AI term, explained in plain English.
A computer chip that does thousands of small calculations at once. First made for video games, it now trains and runs most AI models.
AI Drafted with AI help · Reviewed by Keshav Aggarwal · Updated
In simple words
Imagine one very clever teacher marking 1,000 test papers alone. Now imagine 1,000 students each marking one paper at the same time. The second way is much faster.
A GPU works like the 1,000 students.
The technical side
GPUs have thousands of simple cores built for parallel maths, which suits the matrix multiplications at the heart of neural networks. Nvidia dominates AI GPUs with chips such as the H100 and its successors, along with its CUDA software; AMD and Google’s TPUs are among the alternatives.
GPU supply and price shape what AI companies can build, and renting GPU time is a large part of the cost of every AI answer. See what models cost to use on our LLM API prices page.
GPU kya hai?
GPU ek chip hai jo ek saath hazaaron chhote calculation karti hai. Pehle games ke liye bani thi, ab AI isi pe chalta hai.
Related terms
- InferenceAIThe moment an AI model is used to answer: it takes your input and works out the output. Each answer you get is one inference.
- TransformerAIThe neural network design behind today’s language models. It reads all the words of a text at once and works out which ones matter to each other.
- DePIN (Decentralised Physical Infrastructure)Crypto + AICrypto networks that pay people in tokens for providing real-world equipment, such as GPUs, storage, wireless hotspots or sensors.
