Meta CEO Mark Zuckerberg recently announced that the company is further integrating its AI research team with a business-focused generative AI team, emphasizing their commitment to incorporating the technology into products.

To support this initiative, Meta is expanding its infrastructure and aims to have approximately 350,000 H100 GPUs from Nvidia, along with additional chips from other suppliers, totaling around 600,000 GPUs by the end of the year.

This ambitious plan will position Meta’s system among the largest in the industry, surpassing the scale of competitors like Amazon and Oracle.

While Meta has not disclosed all of its GPU suppliers, it has publicly stated its intention to utilize chips from AMD, in addition to its collaboration with market leader Nvidia. Alongside its previous emphasis on AI research through the FAIR team, Meta has made significant strides in building out its computing capabilities to champion generative AI. One notable example is the establishment of the “GenAI” team, formed in response to the success of OpenAI’s ChatGPT chatbot.

Meta has since introduced various innovative offerings, including a commercial version of its Llama large language model, ad tools capable of generating image backgrounds from text prompts, and the “Meta AI” chatbot accessible through Ray-Ban smart glasses.

Mark Zuckerberg also revealed that Meta is currently training a third version of the Llama model, highlighting the company’s dedication to advancing AI and its alignment with the AR/VR-driven metaverse vision that inspired the name change to Meta in 2021.

As AI continues to evolve, Zuckerberg emphasized the necessity for new devices, such as glasses, to fully engage with this transformative technology.