Graphics Processing Units and AI Revolution

Copyright infringement not intended

Picture Courtesy: The Hindu

Context:

After the discovery of GPU chips by Nvidia in 1999 named GeForce 256 for video games graphics, it became a core infrastructure for the digital economy.

Graphics Processing Unit (GPU):

A Graphics Processing Unit (GPU) is a specialised electronic processor designed to perform a very large number of simple mathematical calculations simultaneously. Unlike a Central Processing Unit (CPU), which is optimised to handle complex tasks and manage system operations, a GPU is built for parallel processing, where the same operation must be repeated across large datasets. Initially developed to render images, videos, and animations for computer graphics, GPUs have become essential for modern workloads such as artificial intelligence, machine learning, scientific simulations, data analytics, and high-performance computing.

Working of a GPU:

The functioning of a GPU is based on a rendering pipeline, which converts 3D scene information into the final image displayed on a screen. This process takes place in the following stages:

Vertex processing

• The GPU processes the vertices (corner points) of 3D objects.
• It calculates the position, rotation, scale, and camera perspective of each object.
• Mathematical operations such as matrix transformations are used to place objects correctly on the screen.

Rasterisation

• The GPU converts geometric shapes (mainly triangles) into pixel coordinates.
• It determines which pixels on the screen are covered by each triangle.
• This stage transforms 3D geometry into pixel candidates.

Fragment (Pixel) shading

• The GPU computes the final colour and appearance of each pixel.
• Effects applied include:
• Textures
• Lighting and brightness
• Shadows
• Reflections and material properties

Frame buffer output

• The final pixel data is stored in a memory area called the frame buffer.
• The display system reads this buffer and renders the image on the screen.
• The GPU performs these operations in parallel across millions of pixels, enabling real-time graphics, high frame rates, and fast computation for graphics as well as AI and scientific workloads.

GPU Location:

A GPU can be located either as a discrete component or as an integrated unit within a system. A discrete GPU is a separate graphics card that plugs into the motherboard through a high-speed interface such as PCIe and includes its own VRAM and cooling system; this configuration is commonly used in gaming computers, workstations, and AI servers. An integrated GPU, on the other hand, is built into the same chip as the CPU and shares the system’s main memory. Integrated GPUs are widely used in laptops, smartphones, and other compact devices where power efficiency and space saving are important.

Source: The Hindu

Comparison between GPU and CPU:

Importance of GPUs for India:

• Foundation for India’s AI led growth: Graphics Processing Units (GPUs) have emerged as a critical enabler of Artificial Intelligence (AI), which is expected to become a major driver of India’s economic transformation. The NITI Aayog in its National Strategy for AI (2018) identified high-performance computing infrastructure as a key requirement for scaling AI applications across sectors such as healthcare, agriculture, education, and governance. According to NASSCOM, AI has the potential to add nearly $500 billion to India’s economy by 2025, highlighting the growing importance of GPU-based computing for training and deploying large AI models.
• Strengthening scientific research and supercomputing: GPUs play a crucial role in accelerating scientific research and high-performance computing in India. Under the National Supercomputing Mission, the government aims to deploy more than 70 supercomputers across academic and research institutions. India’s flagship system, PARAM Siddhi-AI, developed by the Centre for Development of Advanced Computing (C-DAC), uses GPU acceleration and is among the world’s top systems for AI workloads. These GPU-powered systems support critical national functions such as weather forecasting, climate modelling, drug discovery, and space research by the Indian Space Research Organisation.
• Enabling India’s digital economy and startup ecosystem: The growing demand for GPUs is also linked to the rapid expansion of India’s digital economy and startup ecosystem. India currently has over 100,000 recognised startups, according to the Department for Promotion of Industry and Internal Trade, many of which operate in AI-driven sectors such as fintech, health-tech, ed-tech, and deep-tech. Further, the AI startup ecosystem has grown nearly 14 times between 2016 and 2023, as reported by NASSCOM, making access to affordable GPU infrastructure essential for innovation and competitiveness.
• Enhancing global competitiveness: India’s spending on AI is expected to grow at a 25–30% compound annual growth rate, according to the International Data Corporation (IDC), placing it among the fastest-growing AI markets globally. Countries with strong GPU infrastructure gain strategic advantages in emerging areas such as defence AI, cybersecurity, advanced manufacturing, and autonomous systems. Therefore, expanding GPU access and domestic capability is essential for strengthening India’s position in the global technology landscape.

Conclusion:

Graphics Processing Units (GPUs) have become strategic digital infrastructure for India, enabling advancements in artificial intelligence, scientific research, digital governance, and the innovation ecosystem. Expanding access to high-performance computing, strengthening domestic semiconductor capabilities, and promoting energy-efficient data centres will be essential for achieving technological self-reliance, global competitiveness, and sustained digital economic growth.

Source: The Hindu