In a significant development in the field of artificial intelligence, NVIDIA has announced a strategic collaboration with Anyscale, a pioneer in unified computing. This partnership aims to accelerate the development of large language models (LLMs), a crucial component in the AI landscape. The collaboration is set to integrate NVIDIA AI into Anyscale’s open-source Ray platform, the Anyscale Platform, and Anyscale Endpoints, a new service designed to embed LLMs into applications.
NVIDIA and Anyscale partnership
The integration of NVIDIA AI into Anyscale’s platforms is expected to enhance the speed and efficiency of generative AI development. It will also bolster security for production AI, a critical aspect in the current digital age where data breaches and cyber threats are rampant. Developers are given the flexibility to deploy open-source NVIDIA software with Ray or opt for NVIDIA AI Enterprise software on the Anyscale Platform, depending on their specific needs and requirements.
One of the key features of this collaboration is the support of NVIDIA’s TensorRT-LLM, an open-source software, for Anyscale offerings. TensorRT-LLM is designed to run models in parallel over multiple GPUs, offering up to 8x higher performance on NVIDIA H100 Tensor Core GPUs. This performance boost is expected to significantly enhance the efficiency of LLMs, making them more effective and reliable.
Build Large Language Models (LLMs) faster
Other articles you may find of interest on the subject of Large Language Models (LLMs) :
NVIDIA Triton Inference Server software
In addition, the NVIDIA Triton Inference Server software will enable Ray developers to improve efficiency when deploying AI models from various deep learning and machine learning frameworks. This will provide developers with a more streamlined and efficient process for model deployment, reducing the time and resources required.
The NVIDIA NeMo framework, another key component of this collaboration, will allow Ray users to fine-tune and customize LLMs with business data. This customization capability will enable businesses to tailor their AI models to their specific needs, enhancing their effectiveness and relevance.
Anyscale’s platform offers fully managed, enterprise-ready unified computing for building, deploying, and managing scalable AI and Python applications. This collaboration will allow developers to build models on their workstations through NVIDIA AI Workbench and scale them across hybrid or multi-cloud accelerated computing. This scalability is crucial in today’s digital landscape, where businesses need to adapt and scale their operations quickly and efficiently.
The NVIDIA AI integrations with Anyscale are currently in development and are expected to be available by the end of the year. Developers interested in this integration can sign up for updates and a free 90-day evaluation of NVIDIA AI Enterprise. This collaboration between NVIDIA and Anyscale is a significant step forward in the AI industry, promising to bring about faster, more efficient, and more secure AI development.
Source : NVIDIA
Latest Geeky Gadgets Deals
Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, Geeky Gadgets may earn an affiliate commission. Learn about our Disclosure Policy.