Artificial Intelligence (AI) is rapidly transforming industries worldwide, and India is no exception. A significant component of this transformation is the use of Graphics Processing Units (GPUs) and platforms like AI Kosh. This article explores the science behind AI training and the essential role of GPUs in India.
AI Kosh is a revolutionary platform launched by the Indian government. It serves as a repository of non-personal datasets that facilitate AI training. Additionally, it streamlines the process for startups to obtain access to GPUs under the India AI Mission.
GPUs are crucial for AI training due to their exceptional parallel computing capabilities. Unlike traditional CPUs, GPUs can handle multiple operations simultaneously, which significantly speeds up the development of large language models (LLMs). This efficiency is vital for processing extensive datasets required for AI training.
AI training differs markedly from traditional computing. It involves manipulating vast datasets and fine-tuning model parameters through deep learning techniques. Specialized hardware like GPUs is necessary for performing matrix calculations more efficiently than CPUs can.
Foundational AI models are pre-trained on extensive datasets to execute various tasks such as text generation and image recognition. These models serve as the foundation for building customized AI applications tailored to specific requirements.
Currently, India depends on imported GPUs from major companies such as AMD, NVIDIA, and Intel. The cost of GPU power in India ranges from ₹67 to ₹77 per GPU hour, which is relatively lower compared to global data centers that charge approximately $2.25 per GPU hour.
The Indian government is actively engaging in discussions to develop indigenous GPUs, a process expected to take three to four years. This initiative aims to reduce reliance on foreign firms and bolster India’s AI capabilities.
The India AI Mission is designed to enhance AI research and innovation. It provides subsidized access to GPUs for startups and researchers while focusing on creating indigenous AI models that meet India's unique needs.
The government recognizes that AI models trained on global datasets may not accurately reflect India's linguistic and cultural diversity. Consequently, there are plans to develop localized AI models catering to specific Indian languages and sectors, such as agriculture and logistics.
India is also formulating an AI policy framework to ensure ethical AI usage, data privacy, and equitable access to computing resources. Several ministries are launching AI-based services to improve governance and industry applications.
The private sector, including startups like CoRover.ai and BharatGPT, is utilizing government-backed AI infrastructure to create innovative applications. The collaboration between the government and private entities is crucial for advancing AI adoption across various industries.
India aims to reduce AI computing costs by expanding GPU infrastructure and promoting local manufacturing. The subsidized access provided through AI Kosh is set to enhance the accessibility of AI training for startups.
Despite advancements, India faces significant challenges, including dependence on foreign GPU suppliers, high infrastructure costs, and a shortage of skilled professionals. Developing domestic chip manufacturing capabilities remains a long-term goal to overcome these hurdles.
India’s AI policy emphasizes self-reliance, affordability, and ethical development, mirroring global initiatives from the US, EU, and China. However, India is particularly focused on addressing local needs, such as governance and linguistic diversity.
Q1. What is AI Kosh's primary purpose?
Answer: AI Kosh aims to provide non-personal datasets for AI training and facilitate startups' access to GPUs under the India AI Mission.
Q2. Why are GPUs preferred over CPUs for AI training?
Answer: GPUs offer high parallel computing power, allowing for simultaneous processing of multiple tasks, which accelerates AI model development significantly.
Q3. How does India plan to develop indigenous GPUs?
Answer: The government is initiating discussions to create domestic GPUs, expected to take three to four years, reducing dependency on foreign companies.
Q4. What are foundational AI models used for?
Answer: Foundational AI models are pre-trained on large datasets and are essential for executing diverse tasks like text generation and image recognition.
Q5. What is the India AI Mission?
Answer: The India AI Mission promotes AI research by providing subsidized GPU access to startups and focuses on developing localized AI models for India's needs.
Question 1: What is the main function of AI Kosh?
A) To provide personal datasets
B) To offer a library of non-personal datasets
C) To facilitate international AI training
D) To support hardware manufacturing
Correct Answer: B
Question 2: Why are GPUs critical for AI training?
A) They are cheaper than CPUs
B) They can handle parallel processing
C) They are easier to program
D) They require less energy
Correct Answer: B
Kutos : AI Assistant!
