India a strategic market; bullish about our investment commitments: AWS

Bengaluru: India continues to be a very strategic market as far as the overall AWS (Amazon Web Services) framework is concerned, and the company is committed to its investments that are being pushed into the country. Sandeep Dutta, President of India & South Asia, AWS, mentioned that AWS's AI services have reached a $15 billion annual revenue run rate as of Q1 2026 globally, and the momentum is going to be increasing.

"In India, we are seeing a very broad-based demand when it comes to not just cloud, but AI, which is why we continue to deepen our investment profile," he said.

Dutta also mentioned that Amazon plans to invest more than $35 billion across all its businesses in India through 2030, which will build on the $40 billion invested in India so far. Speaking at the AWS Summit here in Bengaluru on Wednesday, Dutta highlighted how AWS has transformed partners' businesses.

Paytm uses AWS to serve over 350 million registered users. Since 2020, AWS has powered 200–250 daily live classes of Physics Wallah reaching 98 per cent of India's pin codes with near-zero downtime. Physics Wallah built Aryabhatta— the first education-focused LLM—delivering personalised learning at 10x lower cost and 5x faster speed than global models.

AWS also showcased its own custom silicon chips. Dutta unveiled the Trainium3 AI chip and Graviton4 processor. This is a 3nm AI chip that provides 2x higher compute performance to 2.52 petaflops (PFLOPs). Graviton is a family of processors designed to deliver for cloud workloads.

Delivering keynote address, Prasad Kalyanaraman, VP, AWS Infrastructure Services said India is one of the six countries in the world that has more than one AWS region- Mumbai and Hyderabad. "These regions are fully featured  ones and they have three availability zones," he said.

Kalyanaraman also mentioned that recently Amazon announced Graviton5, which gives 25 per cent better performance while making the same energy efficiency as previous versions of Graviton.  "It's a great example of how investing in custom silicon allows us to pass on performance gains and cost savings over to you. We've also built a comprehensive, highly scalable, secure AI infrastructure with AWS training designed to offer the best performance for AI workloads with the broadest selection of GPU offerings through our long standing partnership with Nvidia," he said.

Speaking at the Summit,  Swiggy CTO Madhusudhan Rao said soon agents will do all the search for customers and order food according to their preferences. "You have to  open the app, search what you want, browse the menu, add your items and order. All these take nearly eight clicks but now the agent actually does the searching, payments, and comparison of different items. As humans, we can now operate at a higher level," he said, showcasing the platform's demo app 'Sia'.

On April 20, Amazon and Anthropic announced their collaboration with a commitment from the latter to spend more than $100 billion over the next ten years on AWS technologies. This encompasses current and future generations of Trainium (Amazon’s custom silicon) and tens of millions of Graviton cores (Amazon’s widely-adopted CPU chip) to provide superior price performance.

Anthropic will secure up to 5 gigawatts (GW) of capacity to train and power their advanced AI models, including significant Trainium3 capacity expected to come online this year.

Speaking at the event, Anthropic India MD Irina Ghose said India is the second largest base for Claude.ai in the world. "What is also distinguished is the fact that people here are using Claude for development and technical work, which is more than any other place in the world," Ghose said.

Anthropic India MD Irina Ghose speaking at the Summit.

She added that people have found that tasks that they're solving in 15 minutes, were earlier taking more than four hours, a 15x multiplier, and people are really looking at delegating things in an autonomous way, in a friction-free manner.