How is Google Cloud Enhancing AI Capacity in India with IIT Madras?
Synopsis
Key Takeaways
- Indic Arena is a platform for assessing AI models in India.
- Google Cloud is providing cloud credits to support this initiative.
- Local AI hardware capacity is enhanced with Google's AI Hypercomputer architecture.
- Collaboration among various sectors is encouraged for AI development.
- The initiative aims to meet data residency requirements.
New Delhi, Nov 11 (NationPress) In a groundbreaking collaboration, Google Cloud and Google DeepMind unveiled the Indic Arena in partnership with IIT Madras on Tuesday. This public platform empowers Indian users to anonymously assess and rank AI models tailored for the nation’s diverse multilingual environment.
According to a statement, Google Cloud is supplying cloud credits to support this community-oriented initiative operated by the AI4Bharat center at IIT Madras.
Moreover, the tech giant has announced the enhancement of local AI hardware capabilities for Indian customers, which is driven by Google's AI Hypercomputer architecture featuring the latest Trillium TPUs.
This initiative is set to facilitate businesses and public sector entities in training and deploying their most sophisticated Gemini models within India, as stated in the release.
Additionally, Google aims to fulfill data residency and sovereignty requirements with this expansion.
Startups, universities, government organizations, and enterprises are invited to leverage this new AI capacity for Gemini in Vertex AI.
The Indic LLM-Arena serves as a crowd-sourced, human-in-the-loop leaderboard designed to evaluate LLMs based on three crucial elements affecting the Indian experience: language, context, and safety.
“At AI4Bharat, our goal is to develop AI that addresses India’s specific requirements. A vital component of this is establishing a neutral, standardized benchmark to evaluate model performance across our diverse languages,” remarked Mitesh Khapra, associate professor at IIT Madras.
“A model's proficiency in English is meaningless if it cannot comprehend a farmer in rural Maharashtra, gives culturally inappropriate answers to users in Sikkim, or struggles with a Tang-lish query from a student in Tamil Nadu,” clarified a separate statement from IIT Madras.
Earlier this year, Google Cloud made Gemini accessible to regulated Indian customers by introducing Gemini 2.5 Flash with local machine-learning processing capabilities.
IANS
aar/na