Social media giant Meta is embarking on a major shift toward artificial intelligence (AI), with CEO Mark Zuckerberg announcing a significant investment in advanced computer chips from Nvidia to power its AI capabilities. This development is part of Meta’s strategic move to integrate and enhance its AI research and product development efforts.
Build a powerful AI infrastructure with Nvidia’s H100 GPU
Meta’s roadmap includes integrating a whopping 350,000 H100 graphics cards from Nvidia by the end of 2024. Known for their advanced capabilities, these GPUs are essential for Meta’s AI infrastructure, especially the development of artificial general intelligence (AGI). AGI represents a leap forward in AI that aims to achieve human-like cognitive abilities in machines. This significant investment positions Meta at the forefront of AI technology, signifying a deep commitment to pioneering the AI field.
Merging AI research and product development
Meta is undergoing a major internal restructuring to better align AI research and product development. This includes merging the Fundamental AI Research (FAIR) team with its AI product division, GenAI. This integration aims to accelerate the translation of AI research into consumer products, streamline processes, and promote the rapid advancement of AI technology.
Meta’s next-generation Llama 3 large language model
In line with these developments, Meta has begun training the next generation of the Llama 3 large language model. This model is part of Meta’s shift in focus from primarily AI research to integrating AI into social media products, AR/VR hardware, and consumer applications. The Llama 3 model is a further step forward in Meta’s commitment to embed AI deeper into the product ecosystem.
Meta’s AI-based vision for the future
These developments are consistent with Meta’s vision of a metaverse powered by augmented and virtual reality. This vision requires new devices, such as smart glasses, to interact with AI. Zuckerberg’s commitment to this vision is evident in the company’s name change to Meta and its aggressive investments in AI and related technologies.
Comparison with industry giants
To put Meta’s AI initiatives in perspective, Amazon recently announced a system with 100,000 Trainium2 chips, and Oracle released a system with 32,000 Nvidia H100 GPUs. By contrast, Meta’s plans for a total of 600,000 GPUs, including those from Nvidia and other vendors, demonstrate its ambitions to build one of the largest computing systems in the tech industry.
In summary, Meta’s strategic investments and organizational changes reflect a significant shift toward AI, with a particular focus on integrating advanced AI models into its product lineup. This move represents a broader trend in the technology sector toward AI-driven innovation and product development.
Image source: Shutterstock