NVIDIA supports Indian startup Yotta, making the biggest bet on AI in India
Yotta has ordered over 4,000 NVIDIA-manufactured H100 chips and expects to receive approximately 20,000 of these chips in June, marking the largest bet on AI in India to date
Source: Hard AI
Author: Fang Jiayao
Indian startup Yotta, with the support of NVIDIA, is rapidly emerging as a significant force in the Indian AI field.
On March 18th, according to media reports, Yotta has received their first order of over 4,000 NVIDIAs and is expected to receive around 20,000 of the same chips in June, with each chip priced between $30,000 and $40,000. This is India's largest bet on AI to date.
Public information shows that Yotta was co-founded in 2019 by Sunil Gupta, who has decades of experience in the data center industry, and real estate billionaire Niranjan Hiranandani. As a cloud computing service provider, Yotta offers data storage and computing services to clients including Citibank.
In comparison, India's two largest conglomerates, Tata Group and Reliance Industries, also plan to develop AI infrastructure but have not yet ordered NVIDIA's most advanced chips.
Yotta Leading the Rise of Indian AI Technology
With the acceleration of the global AI competition, India, as a rapidly developing market, has seen increasing demand in the field of AI. In September last year, NVIDIA CEO Jensen Huang met with Indian Prime Minister Modi and promised to prioritize orders from Indian data center operators. Huang praised India for its abundant data and talent, stating that India will become one of the world's leading AI markets.
Yotta's CEO Gupta stated that India needs to build AI models tailored to domestic needs to support "national AI" that accommodates local languages and cultural diversity. Therefore, Yotta's strategic goal is to leverage local high-performance computing resources and promote the development of enterprises and research institutions in the field of artificial intelligence within India by providing cost-effective and low-latency services.
Gupta believes that with data centers located within India, the time it takes for data to be transmitted to local users will be shorter, giving Yotta a network latency advantage over foreign cloud service providers, which is crucial for applications that rely on real-time data processing and fast interactions. In addition, Yotta promises to provide the world's cheapest access to NVIDIA AI chips, which will enhance cost competitiveness for Indian businesses and may attract more customers to use Yotta's services.
While Yotta's order volume is not significant on a global scale, for example, tech giants like Microsoft purchase thousands of GPUs, and Meta's CEO Mark Zuckerberg expressed his hope to obtain 350,000 H100s by the end of the year. However, the demand for NVIDIA's H100 GPUs exceeds supply, forcing CEO Jensen Huang to allocate resources accurately under pressure from global corporate giants and world leaders Looking ahead, Gupta believes he can serve clients in India, as well as other clients in Asia and the Middle East. Yotta already has six data centers in four cities in India, with a new data center in the northeast of India set to open soon. Through this model, Yotta hopes to build a relatively independent and robust local AI ecosystem.
Gupta stated, "I am ambitious and willing to bet on the future of artificial intelligence."
Gupta's ambition aligns with the explosive demand growth for artificial intelligence. According to a report by Spherical Insights & Consulting, the global AI market is expected to grow significantly over the next decade, from $168.5 billion in 2022 to over $2 trillion in 2032.
Analyst Stacy Rasgon from the investment bank Bernstein likened this situation to a "gold rush." She pointed out that we are still in the early stages of AI development, with companies having a strong demand for AI technologies to the extent that they can hardly find enough related products and services