Sridhar Vembu says India must prioritise tech sovereignty over spending billions on GPUs

Sridhar Vembu wants India to prioritise infrastructure and foundational technologies over heavy spending on GPUs. His core idea is that durable prosperity depends on capabilities that support the whole economy, not only the latest AI systems. This matters because dependence on foreign technology can create vulnerabilities during geopolitical tensions.

He gave a striking example: if he earned $10 billion as a government, he would spend it on Bangalore and Chennai infrastructure rather than GPUs. He also wants technology investment to reach agriculture and manufacturing. Better infrastructure and technical know-how could support food production, industrial capacity, and wider economic growth.

The current debate is shaped by excitement about AI and anxiety over possible technology job losses. Vembu says India should focus on foundational and upward-building priorities. That approach would aim to create prosperity that remains in India and make the country less exposed to disruptions in access to critical foreign technologies.

A GPU, or graphics processing unit, is a processor designed to perform many calculations at the same time. Although originally developed for graphics, this parallel structure makes GPUs valuable for artificial-intelligence computing. AI systems repeatedly process large amounts of numerical data, so parallel calculation can accelerate that work.

For example, training an AI model involves adjusting many numerical values across repeated calculations. GPUs can handle numerous parts of those calculations together. They are therefore especially useful for building and running advanced AI systems, which helps explain the current excitement around purchasing them.

The article does not explain how GPUs work technically, but it presents them as a major focus of current AI investment. Vembu argues that India should not make GPU purchases its main priority. Instead, he favours infrastructure and foundational capabilities that can support AI, agriculture, manufacturing, and broader economic development.

Vembu illustrates his argument with a hypothetical $10 billion government windfall. He says that, if he earned that amount as a government, he would spend it on infrastructure in Bangalore and Chennai. He would not spend the money on GPUs immediately. The large figure highlights the scale of investment he believes should move toward broad public foundations.

His wider concern is that technology discussions often revolve around software and AI while overlooking most people. Vembu says approximately 97% of the population does not care about software technology and remains on the consuming side. This suggests that technology policy should serve people and sectors beyond the technology industry.

The article places these comments amid excitement and anxiety about AI, including fears of job losses among technology professionals. Vembu therefore calls for investment in agriculture, manufacturing, infrastructure, and foundational technologies, rather than concentrating resources narrowly on AI hardware.

If geopolitical tensions disrupt access to foreign-made GPUs and other critical technologies, India could face serious technology and economic vulnerabilities. The article does not specify a particular failure or forecast a precise outcome. It does state that dependence on foreign countries could leave India exposed when access to important technologies is interrupted.

The key mechanism is technological dependence. If essential hardware, foundational tools, or technical capabilities come from abroad, disruptions could affect India’s ability to develop and use advanced systems. This concern extends beyond GPUs because Vembu calls for sovereignty in foundational technologies across the economy.

The current reality is growing excitement around AI alongside anxiety about its effects, including possible job losses. Vembu’s proposed response is to build domestic infrastructure and know-how. He believes India should prioritise foundational and upward-building capabilities so prosperity stays with the country and access to critical technologies is more secure.

Vembu wants India to invest in infrastructure and foundational technologies instead of focusing mainly on GPUs. He specifically mentions infrastructure in Bangalore and Chennai. He also says technology investment should extend beyond software into agriculture and manufacturing. These foundations matter because they can support many sectors, not just AI companies or technology professionals.

The key mechanism is applying technical know-how to real economic needs. In agriculture, better knowledge and tools could help India grow more food. In manufacturing, stronger foundational capabilities could support production and economic activity. Vembu’s argument is that technology should reach the consuming majority and the sectors that produce essential goods.

The article does not list specific projects, devices, or programmes. It does show the direction Vembu favours: build infrastructure, develop foundational technologies, and reduce foreign dependence. He believes these priorities can create durable prosperity that remains with India, especially amid geopolitical uncertainty.

Vembu points to several countries to show how technology can support food production and economic prosperity. He names China, South Korea, and Taiwan as countries that have leveraged technology to boost food production and drive economic growth. His examples broaden the discussion beyond software and AI.

The key mechanism is practical know-how applied to physical sectors. Vembu says India needs much more know-how to grow more food. He also highlights Israel, saying it is able to grow food in the desert. The example illustrates how technology and expertise can overcome difficult production conditions and expand food-growing capacity.

The article does not identify particular technologies or explain the methods used in each country. It presents their results as examples for India to consider. Vembu’s wider recommendation is to direct investment toward agriculture, manufacturing, infrastructure, and foundational technologies so prosperity becomes more durable and less dependent on foreign countries.

Technological sovereignty means having enough domestic capability to develop, maintain, and use critical technologies without excessive dependence on foreign countries. In Vembu’s argument, it is part of building prosperity that stays with India. It matters because geopolitical tensions could disrupt access to technologies that a country relies on.

Software alone cannot provide that independence. Hardware such as GPUs, manufacturing capacity, physical infrastructure, and technical know-how are also needed to create and apply technology. These capabilities can support AI, agriculture, manufacturing, and food production. They also help technology serve people outside the software industry.

The article does not provide a detailed national blueprint or name specific hardware programmes. It does set a clear direction: prioritise foundational and upward-building capabilities. Vembu contrasts this approach with spending $10 billion on GPUs right now. His goal is durable prosperity and reduced vulnerability to foreign supply disruptions.