Meta to Purchase Up to 1.3 Million Nvidia GPUs for Artificial Intelligence

Mark Zuckerberg’s company is preparing to compete with Google and OpenAI with a record investment that includes 1.3 million GPUs and the construction of a gigantic data center.

The advancement of artificial intelligence (AI) has transformed the way companies compete in the tech market. Since OpenAI introduced ChatGPT to the world in 2022, giants like Google, Meta, and Anthropic have begun a race to develop the most advanced AI models. In this context, Meta has announced that it will allocate an unprecedented investment of $65 billion in 2025 to strengthen its artificial intelligence infrastructure and consolidate its position in this field.

An ambitious goal: 1.3 million GPUs

One of Meta’s main targets is to acquire over 1.3 million GPUs from NVIDIA, a clear leader in hardware for artificial intelligence. According to internal sources, the company aims to achieve computing power equivalent to 1 gigawatt (1 billion watts), which would represent a significant leap in its infrastructure. This move aligns with Meta’s increasing focus on Llama, its AI-based language model that directly competes with those from OpenAI and Google.

Each NVIDIA H100 GPU, the model chosen by Meta, costs about $30,000, putting the colossal cost of this strategy into perspective. It’s worth noting that in 2024, Meta had already acquired 350,000 of these GPUs, a figure that was impressive at the time, but now pales in comparison to this year’s ambitious plan.

The construction of a colossal data center

Meta’s investment is not limited to hardware acquisition. The company has also announced the construction of a massive data center, which is expected to be one of the largest in the world. According to statements from Mark Zuckerberg, this center will be so large that it could cover an area similar to a significant part of Manhattan. Its energy consumption will exceed 2 gigawatts, reflecting the project’s titanic scale.

meta Geothermal Energy Data Centers

In addition to hardware, Meta plans to hire thousands of specialized employees in artificial intelligence to ensure the development and efficient operation of its infrastructure. These hires are part of a broader plan to ensure that Meta has the necessary human and technical capabilities to compete with rivals like OpenAI and Google, who have already shown significant advancements in AI with ChatGPT and Gemini, respectively.

An ever-evolving market

Since the launch of ChatGPT, the landscape of artificial intelligence has changed radically. Google, with its Gemini model, and Anthropic, with Claude, have demonstrated impressive advancements in the development of generative AI, while Elon Musk has also joined this race with his xAI initiative. Musk already has a supercomputer equipped with over 100,000 NVIDIA GPUs, but Meta’s plans aim to far surpass this number.

Meta is well aware of these competitive dynamics. Its bet on Llama, a model that promises to rival the best generative AI systems, has become the central axis of its strategy. However, the company’s goal is not only to develop the most powerful model but also to ensure that it can operate on a global scale, with millions of simultaneous users.

The highest spending in Meta’s history

The investment announced by Meta for 2025 represents an increase of nearly 50% compared to the previous year, even exceeding the most optimistic estimates by analysts, who had projected spending of approximately $50.25 billion. This increase reflects Meta’s urgency to close the gap with Google and OpenAI, especially in a year that Zuckerberg himself has described as “decisive for artificial intelligence.”

In addition to hardware and data centers, Meta also plans to invest in software and tools for AI, which include advanced customization capabilities and greater integration into its platforms, such as Instagram, WhatsApp, and Facebook.

A future defined by artificial intelligence

Meta’s commitment to artificial intelligence is not only a response to current competition but also a long-term vision where AI will play a central role in all aspects of technology and society. From training models to implementation in everyday applications, such as virtual assistants or data analysis systems, AI is destined to be an essential component of our digital lives.

Mark Zuckerberg has made it clear that this investment is just the beginning of a new phase for Meta. “2025 is a critical year for artificial intelligence. We cannot afford to fall behind at this pivotal moment,” stated the CEO. With this statement, Meta underscores its intention to be not just a competitor, but a leader in the field of artificial intelligence in the coming years.

In a world where technology is advancing at breakneck speed, Meta is betting big to stay at the forefront. Only time will tell if this colossal investment can fulfill its ambitious goals and redefine the future of artificial intelligence.

via: Reuters and El chapuzas informático

Scroll to Top