This announcement underlines the deepening ties between Meta and Nvidia. Nvidia plays a crucial role, supplying Meta with GPUs essential for training these expansive AI models.
This announcement underlines the deepening ties between Meta and Nvidia. Nvidia plays a crucial role, supplying Meta with GPUs essential for training these expansive AI models.
Unlike competitors such as OpenAI and Google, which monetize their AI models, Meta continues to shun the creation of a parallel enterprise business. Instead, it has opted to collaborate with numerous tech companies, allowing them to integrate Llama 3.1 into their cloud platforms and provide additional security and management tools.
A spokesperson highlighted that while the Llama partnerships yield some revenue, the primary aim is to foster a talent-rich, competitive edge and reduce overall computing costs by leveraging the open-source model.
Notably, Meta does not operate a business-facing cloud like some of its counterparts but remains one of Nvidia’s top clients. The robust partnership is evident as Meta utilizes Nvidia's cutting-edge H100 GPUs to train its models, crucial for internal operations such as targeted advertising.
The open-source nature of Meta’s AI projects serves to broaden the potential application of Nvidia’s hardware, sustaining demand without the need for licensing fees. This model also aligns with Meta's strategy of cultivating an ecosystem where developers create applications using freely available resources, enhancing both accessibility and innovation within the tech community.
In a recent blog post, Zuckerberg emphasized Meta’s unique approach to fostering broad partnerships, enabling a diverse range of functionalities for client companies.
The flagship model of the Llama 3.1 series, named Llama 3.1 405B, incorporates 405 billion parameters, significantly enhancing its ability to manage complex tasks such as understanding extensive texts, solving intricate mathematical problems, and generating synthetic data to augment smaller models.
Smaller versions, Llama 3.1 8B and Llama 3.1 70B, are also released, offering enhancements over their predecessors and supporting applications like chatbots and coding assistants.
U.S.-based users of WhatsApp and visitors to Meta.AI will soon experience firsthand the capabilities of the new Llama 3.1 through interactions with Meta's digital assistant, capable of addressing complex queries and coding issues. Users can choose between the high-capacity Llama 3.1 model or a smaller, faster version tailored to their needs.
Sara Montes de Oca is the Editor in Chief of TechEchelon. Previously a correspondent and producer in Washington, D.C., covering business, finance, and politics.
More from Sara →Stifel upgraded Shopify to buy on Friday and raised its price target to $150 from $110, arguing that agentic AI commerce will accelerate the platform's share gains even as the technology's financial impact remains in early stages.
Meta launched Muse Spark 1.1 with aggressive developer pricing, OpenAI broadly released its GPT-5.6 series with a 54% efficiency claim, and Anthropic introduced a usage-analytics feature for Claude — all on the same day.
Meta Platforms announced plans to build a 1 gigawatt data center in Alberta, Canada — its first major facility in the country — at an estimated cost of $9 billion, as the company races to expand AI infrastructure capacity.
Written for readers who already know the basics — markets, AI, and the policy decisions that shape both.