In a time when every enterprise and nation is trying to build their own sovereign AI, Nvidia has introduced a new offering under its Nvidia AI Foundry using which enterprises and nations can now build their own AI supermodel based on Llama 3.1 tailored to reflect their unique business and culture.
Previously, NVIDIA AI Foundry offered NVIDIA-created AI models like Nemotron and Edify, along with popular open foundation models, enabling users to train and deploy custom AI supermodels. Now with its partnership with Meta, it extends its capabilities to include Meta’s latest open-source model, Llama 3.1.
Nvidia Founder and CEO Jensen Huang emphasized the significance of Llama 3.1 integration with Nvidia AI Foundry, stating
Meta’s openly available Llama 3.1 models mark a pivotal moment for the adoption of generative AI within the world’s enterprises
What is Meta’s Llama 3.1?
Llama 3.1 is the latest collection of open-source AI large language models comprising both pretrained and instruction-tuned text-in/text-out generative AI models available in 8B, 70B, and 405B parameter sizes from Meta. The instruction-tuned Llama 3.1-405B stands out as the largest and most powerful open-source AI model.
When asked from Meta Founder and CEO Mark Zuckerberg about their latest open source AI offering , he highlighted the importance of Llama 3.1 for open-source AI, stating:
The new Llama 3.1 models are a super-important step for open source AI.
What makes this Meta and Nvidia AI Foundry so special is that , all these large language models under Llama 3.1 has been trained over 16,000 NVIDIA H100 Tensor Core GPUs and are optimized for NVIDIA accelerated computing and software — in the data center, in the cloud and locally on workstations with NVIDIA RTX™ GPUs or PCs with GeForce RTX GPUs.
//www.nvidia.com/en-us/design-visualization/desktop-graphics/" target="_blank" rel="noopener noreferrer">NVIDIA RTX™ GPUs or PCs with GeForce RTX GPUs.For in-depth details on Meta’s Llama 3.1, refer to IBM’s blog.
