Graphics Processing Unit (GPU)
Kartavya Desk Staff
Source: TH
Subject: Science and Technology
Context: Graphics Processing Units (GPUs) are in focus as they have become the backbone of modern AI systems, cloud computing, and high-performance digital infrastructure.
About Graphics Processing Unit (GPU):
What is a GPU?
• A Graphics Processing Unit (GPU) is a specialised computer processor designed to perform many simple calculations simultaneously, making it ideal for parallel processing tasks.
• Unlike CPUs, which handle fewer complex tasks, GPUs excel at repetitive, data-intensive computations.
Origin:
• The term GPU gained prominence in 1999, when Nvidia launched the GeForce 256, marketed as the world’s first GPU.
Aim: The primary aim of a GPU is to handle embarrassingly parallel workloads—tasks that can be broken down into thousands of smaller, independent calculations performed simultaneously.
How it Works?
A GPU works through a process called a rendering pipeline (or a compute pipeline for non-graphics tasks):
• Vertex Processing: It calculates the position of 3D objects on a 2D screen using matrix mathematics.
• Rasterization: It converts these geometric shapes into pixels (fragments).
• Shading: It determines the color, lighting, and texture of each pixel simultaneously across thousands of cores.
• Output: The final image is written to VRAM (Video RAM) and sent to the monitor.
In modern AI, the GPU skips the visual steps and uses its cores to perform massive matrix multiplications, which are the mathematical foundation of neural networks.
Key Features:
• Parallel Architecture: Contains hundreds or thousands of small, specialized cores (e.g., CUDA cores or Tensor cores).
• High Memory Bandwidth: Uses specialized memory like GDDR6X or HBM3 (High Bandwidth Memory) to move massive amounts of data quickly.
• Programmability: Through platforms like Nvidia CUDA or OpenCL, developers can use GPUs for non-graphics tasks (GPGPU).
• Energy Density: High-end GPUs in 2026 can consume over 1000W per device, requiring advanced liquid cooling in data centres.
Applications:
• Artificial Intelligence: Training and running Large Language Models (LLMs) like GPT-4 or Gemini.
• Gaming & VR: Real-time ray tracing and high-frame-rate 4K/8K rendering.
• Scientific Simulation: Weather modelling, molecular dynamics for drug discovery, and genomic sequencing.
• Professional Visualization: 3D CAD modelling, video editing, and digital twins for industrial AI factories.
• Blockchain: Handling complex Proof of Work hashes for cryptocurrency mining.