Minister Ashwini Vaishnaw has announced at Utkarsh Odisha Conclave that India is preparing to create a unique artificial intelligence (GENAI) model. As a result, South Asian countries are arranged with global AI leaders such as Chatgpt and DeepSeek.
This project is an Indian AI Computing Facility with over 18,000 graphics processing unit (GPU) to support the creation of a large language model (LLM) designed to meet the specific needs of the country. It is supported by.
LLMS works as a Genai core. Global and general -purpose LLMS supports a variety of use cases and has a wide range of applicability, but there are various reasons and issues related to the development of models according to local languages.
“The whole world believes that India will be a major semiconductor destination. We will soon get its own LLM,” said Weissuno.
“As 18,000 GPUs are enforced, India is going smoothly to build a unique Language, Economic, and Social requirements in its own country,” he added. I did it.
The government is reportedly chosen 18 proposals to receive support in the form of computing in frustration, data, and funds for the development of AI applications in areas such as agriculture and climate change. According to Vaishnaw, India covers 40 % of these proposal costs.
“The basic model made in India will be able to compete with the best in the world,” said Vaishuno.
“We believe that at least six major developers can develop AI models in 6-8 months due to the outside restrictions can develop 4-6 months with more optimistic estimates. This is the most important component for creating a robust AI ecosystem, “said Vaishnaw.
An important factor in Indian missions is that the government focuses on creating intensive computing resources. Vaishnaw has shared the fact that facilities have surpassed their first goals, have earned more than 18,000 GPUs, far exceeding the target of 10,000. This includes 12,896 NVIDIA (NASDAQ: NVDA) H100 GPU and 1,480 NVIDIA H200 GPU, which contains some of the most advanced AI processors in the market. Approximately 10,000 GPUs are already working, and the remaining GPUs are set to be deployed over time.
Vaishnaw states that the AI calculation service is offered at a average discount of 42 % compared to market prices.
India will also introduce its own basic AI models aimed at working on the country’s linguistic and cultural diversity while reducing the bias of datasets. Vaishnaw has revealed that it will begin with the expected 4-8 month development period in search of proposals to develop AI models.
Who controls the AI sector?
India’s announcement occurs when the United States and China are engaged in fierce competition for dominance in AI. This movement matches the Startups of Deepseek AI in China, creating a headline with that break -through, and challenging a giant in the United States like Openai.
Deepseek, a Chinese AI startup, is a Chinese startup just a year ago, and by introducing AI models that provide the same performance as the world’s top chatbots in only a part of the development costs, Silicon Valley. So I caused both excitement and concerns. On January 27, the global high -tech stock was reduced as the topic surrounding DeepSeek’s innovation grew. Investors have encouraged them to consider the potential impact on US competitors and AI hardware providers, such as NVIDIA CORP. The rapid decrease in high -tech shares reduces the market value of nearly $ 600 billion.
NVIDIA’s shares struggled with 17 % of the number of 17 %, the largest decrease in market history. Other technology sector shares, including semiconductors, electricity, and AI -related infrastructure sectors, have been collective losses with a market value of over $ 1 trillion.
“The release of Deepseek AI from a Chinese company should be a call for an industry that needs to focus on lasers by competing to win, because there are the greatest scientists in the world. Trump said. Blood bus in Wall Street.
This statement is a few days after Sam Altman, the CEO of Masayoshi, the son of Masayoshi of Trump, Oracle, Larry Ellison, and SFTBY, announced the Star Gate, a venture of the AI data center. I am. Stargate is a shared initiative led by Oracle, SoftBank, and Openai, aiming to invest $ 100 billion to $ 500 billion in the development of US AI infrastructure.
On the other hand, the Mukeshu Ambani Rilean Group, a billionaire, has purchased NVIDIA AI semiconductors to establish a data center in India. The data center is expected to make the total capacity of the world’s largest data center, 3 gigawatts, depending on the capacity.
Indian server Deepseek
Vaishnaw has announced that DeepSeek will soon be hosted on an Indian server to work on privacy concerns.
“This is an open source model. Like the open source LLAMA, it can be hosted on an Indian server. Data privacy issues about DeepSeek can be dealt with by hosting the open source model on the Indian server. “Vaishnaw explained.
In 2024, India approved about $ 1.24 billion in Indian missions. It aims to boost the AI ecosystem, innovation, and entrepreneur spirit of the country.
“Some people have questioned the investment amount committed to the (Indian mission). Did you see what DeepSeek did? A very powerful model of $ 5.5 million. For, “said Vaishuno.
India is planning to manufacture its own “India” chips by the end of 2025, and plans to investigate the possibility of having its own GPU in 3-5 years. Manufacturing unique chips is an important movement to accelerate semiconductor ambitions, but Indian GPUs do not depend on foreign suppliers alone, but support the development of AI infrastructure. Useful for.
In order for artificial intelligence (AI) to work correctly in the law and prosper in the face of an increase in issues, it is necessary to integrate an enterprise blockchain system that guarantees the quality and ownership of data input. Data. Check out the coingeek reports on this new technology and learn more why Enterprise blockchain becomes AI backbone.
WATH: Blockchain, IPv6, AI, 5G will open the new Internet
https://www.youtube.com/watch?v=SVSEO2UNJEM Title = “YouTube Video Player” FrameBorder = “0” APLAS = “Accelerometer; Clipboard-Write; Picture; Web-Share “ReferrerPolicy =” Strict-Origin-WHEN-WHEN-CROSS- Origin “approakfullscreen =” “>