Snowflake (NYSE: SNOW), the AI data cloud company, today at Snowflake Summit 2024 announced a new collaboration with NVIDIA that customers and partners can leverage to build customized AI data applications on Snowflake, powered by NVIDIA AI.
With this latest collaboration, Snowflake has adopted NVIDIA AI Enterprise software and integrated its NeMo Retriever microservice into Snowflake Cortex AI, Snowflake's fully managed large-scale language model (LLM) and vector search service, enabling organizations to seamlessly connect custom models to a variety of business data and deliver highly accurate responses. In addition, Snowflake Arctic, the most open enterprise-grade LLM, is now fully supported by NVIDIA TensorRT-LLM software, providing users with highly optimized performance. Arctic is also now available as an NVIDIA NIM inference microservice, making Arctic's efficient intelligence accessible to more developers.
“As businesses look for ways to unlock more of the power of AI across their teams, there is an increasing need to apply data to drive customization. Snowflake and NVIDIA's collaboration enables organizations to quickly create bespoke AI solutions specific to their use cases, helping companies across industries realize the potential of enterprise AI.”
“Combining NVIDIA's full-stack accelerated computing and software with the cutting-edge AI capabilities of Snowflake's Cortex AI is groundbreaking,” said Sridhar Ramaswamy, CEO of Snowflake. “Together, we are ushering in a new era of AI where customers across every industry and skill level can easily, efficiently, and reliably build custom AI applications on their enterprise data.”
“Data is the essential raw material for the AI industrial revolution,” said Jensen Huang, founder and CEO of NVIDIA. “Together, NVIDIA and Snowflake are helping companies refine their own business data and transform it into valuable generative AI.”
Snowflake Cortex AI + NVIDIA AI Enterprise Software
Snowflake and NVIDIA are collaborating to integrate key technologies from the NVIDIA AI Enterprise software platform, such as NeMo Retriever, into Cortex AI, enabling business users to efficiently build and leverage custom AI-powered applications that make the most of their AI investments.
NVIDIA AI Enterprise software capabilities delivered with Cortex AI include:
- NVIDIA NeMo Retriever: Provides high accuracy and powerful performance for information retrieval for businesses building search-enhanced generative-based AI applications within Cortex AI.
- NVIDIA Triton Inference Server: Provides the ability to deploy, run and scale AI inference for any application on any platform.
Additionally, NVIDIA NIM inference microservices (part of NVIDIA AI Enterprise, a set of pre-built AI containers) can be deployed directly within Snowflake as a native app with Snowpark Container Services, enabling organizations to easily deploy a set of foundational models directly within Snowflake.
Quantiphi, an AI-first digital engineering company and an “Elite” level partner of both Snowflake and NVIDIA, is one of many AI providers using Snowpark Container Services to build Snowflake native apps. These apps run within customers' Snowflake accounts, ensuring data protection while accelerating time to value. Quantiphi's native apps, baioniq™ (a generative AI platform for knowledge worker productivity) and Dociphi (an AI-driven intelligent document processing platform for banking, financial services, and insurance), target specific business personas to accelerate industry use cases and daily operations. Both Dociphi and baioniq are developed using the NVIDIA NeMo framework and are offered on the Snowflake Marketplace, allowing users to deploy them without leaving their Snowflake environment.
Expanded support for Snowflake Arctic
The state-of-the-art Snowflake Arctic LLM, released in April 2024 and trained on NVIDIA H100 Tensor Core GPUs, will be available as an NVIDIA NIM, allowing users to get started with Arctic in seconds. The NVIDIA-hosted Arctic NIM is published in the NVIDIA API catalog for developers to access using free credits, and is available as a downloadable NIM, giving users even more choice to deploy the most open enterprise LLM available on their preferred infrastructure.
Earlier this year, Snowflake and NVIDIA announced the expansion of their initial collaboration to deliver a single, unified AI infrastructure and computing platform on the AI Data Cloud. Today's announcement represents an important step forward in Snowflake and NVIDIA's shared mission to help customers succeed in their AI initiatives.
read more…
