MeitY Eases AI Compute Procurement Norms to Aid Start-ups

Source: This post on MeitY Eases AI Compute Procurement Norms to Aid Start-ups has been created based on the article “To accommodate start-ups, MeitY relaxes AI compute procurement norms” published in Indian Express on 9^th October 2024.

Why in news?

The Ministry of Electronics and Information Technology (MeitY) has made adjustments to its AI compute procurement norms to foster inclusivity, especially for smaller companies.

Key Relaxations

1. Objective: The relaxations aim to empower Indian start-ups to participate in AI infrastructure development, fostering innovation and supporting India’s AI growth, especially in critical sectors like healthcare, agriculture, and governance.

2. This initiative is part of the Rs 10,370 crore India AI Mission, which aims to establish a computing infrastructure with over 10,000 GPUs.

3. It also seeks to support the development of foundational models with capacities exceeding 100 billion parameters, trained on datasets that encompass key Indian languages.

4. These models will address priority sectors such as healthcare, agriculture, and governance.

5. The goal is to create an infrastructure within the country that start-ups can utilize to develop AI systems.

6. Annual Turnover Requirement:

 For primary bidders: Lowered from Rs 100 crore to Rs 50 crore.

 For non-primary consortium members: Halved from Rs 50 crore to Rs 25 crore.

7. Compute Capacity Requirements:

FP16 performance threshold reduced from 300 TFLOPS to 150 TFLOPS.

AI compute memory requirement lowered from 40 GB to 24 GB.

8. Experience Criteria: Bidders must now disclose AI services provided in the last three financial years (2020-2023), with a minimum billing of Rs 10 lakh. Extra points are awarded based on the number of clients serviced.

9. Local Component Sourcing: In line with ‘Make in India’ guidelines, components for cloud services must be sourced from Class I (50%+ domestic value) or Class II (20-50% domestic value) local suppliers.

10. Implementation Approach: Public-Private Partnership where 50% viability gap funding is offered to private entities to develop the AI infrastructure, with Rs 4,564 crore specifically allocated for building computing capacity.

UPSC Syllabus: Indian Economy