Five years ago, NVIDIA’s CEO Jensen Huang delivered the first DGX AI supercomputer to OpenAI, which was just a startup at the time. In the five years since, OpenAI has made great strides in AI research and development.
This raises the question of how long it might take major Indian companies like Reliance or Tata to achieve a similar level of AI expertise and impact as OpenAI has in this relatively short timeframe.
While it's difficult to predict exactly, Indian tech giants have access to world-class engineering talent and sizable resources. With concentrated investment and focus in developing internal AI capabilities, we could expect to see companies like Reliance and Tata make meaningful advancements in AI over a 5-10 year timeline, though likely not yet matching OpenAI's accomplishments thus far.
The pace of progress will depend on factors like:
- Research priorities
- Attracting top AI researchers
- Acquiring or building the necessary computing infrastructure
The Scale of OpenAI's AI Infrastructure
In an exclusive interview, NVIDIA CEO Jensen Huang stated that OpenAI currently has over 10,000 GPUs, which is equivalent to around 40 AI supercomputers.
Huang explained that each AI supercomputer rack consists of 256 GPUs. This large GPU-powered infrastructure allows OpenAI to train complex AI models that require immense computational power.
The scale of OpenAI's GPU resources highlights their position as an industry leader in AI research, with the hardware capabilities to develop and experiment with cutting-edge AI systems. Access to such substantial GPU-accelerated computing is a key factor empowering OpenAI's rapid progress in AI capabilities like language models and robotics.
Estimating India's Path to Matching OpenAI's Resources
Based on OpenAI's GPU infrastructure, Jensen Huang estimated that India would need on the order of 100,000 GPUs, or less than 400 AI supercomputers, in order to build similar AI capabilities and infrastructure.
By quickly doing the math, Huang concluded that the number of GPUs required for India to match OpenAI's level of AI infrastructure and research would be in the tens of thousands range. This would equate to below 400 racks of GPU servers, with each rack containing around 256 GPU chips.
While a massive undertaking, this level of investment in GPU-powered supercomputing could allow Indian companies or research institutions to train and experiment with the most advanced AI models at scales comparable to OpenAI.
However, accumulating this order of GPU-enabled systems would still represent a formidable challenge for building homegrown AI that can compete at the cutting edge globally.
NVIDIA Providing India the Tools for AI Innovation
Jensen Huang is very bullish on the future of AI in India. He stated that where it takes US companies like OpenAI months to train large AI models like GPT, India could achieve similar results in just weeks with the new GPU systems NVIDIA will provide.
Huang confirmed that NVIDIA is bringing "the fastest computers in the world" to India soon, faster than anything currently available and highly cost-effective. By end of next year, India will have AI supercomputers over 50-100x faster, dramatically reducing the cost to train foundational AI models.
To start with, Reliance and Tata will have access to NVIDIA's DGX Cloud service, allowing any business to utilize an AI supercomputer via a web browser without complex on-site infrastructure. "You can build a large language model, like a ChatGPT for $10-$20 million," said Jensen, explaining how NVIDIA has made advanced AI development affordable by leveraging cloud infrastructure.
Huang believes NVIDIA's new GH200 Grace Hopper Superchip and DGX Cloud AI supercomputing service will further accelerate India's AI capabilities.
India's Advantages for Homegrown AI
Jensen highlighted India's competitive advantages in AI:
- Abundant data, across diverse languages and dialects, that doesn't need to be exported to Western companies
- Strong homegrown computer science talent, with India producing more CS grads than any country
- Existing education infrastructure like the IITs for developing AI talent
However, he noted India currently lacks infrastructure specifically for AI research and development. But with NVIDIA's incoming supercomputers, Jensen believes this gap is also being addressed.
In his view, India has all the elements needed to build and deploy AI internally: data, human capital, and with NVIDIA's help the infrastructure.
He compared this to electrifying the nation with power plants and steam engines - now India can produce intelligence by investing in AI infrastructure.
Jensen stressed that India doesn't need to rely on external tech companies, but has the resources and skills to create world-class homegrown AI and foundational models like LLMs, powered by domestic data.
With strategic investment in GPU supercomputing, India can unlock its full AI potential.
NVIDIA Partnerships Driving India's AI Revolution
According to Jensen Huang, India's AI revolution is beginning through partnerships between NVIDIA, Reliance, and Tata:
- Reliance will use NVIDIA infrastructure to create customer-focused AI services for its 450 million Jio users, and provide efficient AI resources for Indian startups and researchers.
- Tata will leverage NVIDIA's expertise and systems to build enterprise generative AI applications for departments like legal, HR, and sales. This expands Tata's cloud services to meet demand from generative AI startups and large language model processing.
- TCS will use the AI infrastructure to develop and run generative AI apps. They will also upskill their 600,000 employees through the partnership.
- Tata Communications' global network and NVIDIA's AI cloud will enable high-speed data transfer, bringing AI cloud access to every enterprise.
With Reliance serving consumers, Tata servicing enterprises, and TCS readying the workforce, NVIDIA's infrastructure and knowledge transfer can catalyze India's AI revolution.
Huang sees this as just the beginning, with huge potential for homegrown AI innovation and adoption in India.
Hey, join our AI SubReddit, Facebook Community, Discord Channel, and Email Newsletter, where we share the latest AI research news, awesome AI projects, AI guides/tutorial, Best AI tools, and more.
Subscribe to our daily newsletter to receive the top headlines and essential stories delivered straight to your inbox. If you have any questions or comments, please contact us. Your feedback is important to us.
