Snowflake Users Set to Receive RelationalAI's Knowledge Graph Coprocessor

An artificial intelligence-powered tool is reportedly set to grace Snowflake users as RelationalAI announces its Knowledge Graph Coprocessor is coming to the cloud-based platform.

RelationalAI, a startup that uses artificial intelligence for relational data, has announced the general release of its Knowledge Graph Coprocessor on the Snowflake Data Cloud. The solution, initially revealed last year, allows Snowflake clients to create knowledge graphs and access powerful AI and analytics capabilities without transferring data outside their Snowflake environment.

According to Molham Aref, cofounder and CEO of RelationalAI, this is a significant victory for the client since, typically, to perform prescriptive analytics, predictive analytics, graph analytics, or rule-based reasoning, users must extract data from Snowflake and place it in a point solution.

The RelationalAI Knowledge Graph Coprocessor is designed to apply AI directly to structured corporate data in relational databases such as Snowflake. This contrasts with most machine learning and artificial intelligence techniques, which focus on unstructured data such as text, photos, and audio.

Snowflake Nvidia Partnership

Snowflake's latest launch of an AI-focused solution comes only a few days after the cloud platform announced a new relationship with Nvidia at Snowflake Summit 2024. The collaboration will enable clients and partners to create bespoke AI data applications in Snowflake, which Nvidia AI powers.

In this cooperation, Snowflake has integrated Nvidia AI Enterprise software, combining NeMo Retriever microservices into Snowflake Cortex AI, which enables the smooth coupling of bespoke models to different business data, boosting the delivery of exact solutions.

Furthermore, Snowflake Arctic, an enterprise-grade large language model (LLM), is now completely compatible with Nvidia TensorRT-LLM software, resulting in significantly optimized performance. Arctic is also available as a Nvidia NIM inference microservice, which makes it more accessible to developers.

Snowflake CEO Sridhar Ramaswamy stressed the revolutionary power of merging Nvidia's accelerated computers and software with Snowflake's sophisticated AI capabilities, calling it game-changing.

The cooperation intends to usher in a new AI era by allowing customers of all skill levels and sectors to easily and confidently create unique AI solutions.

Jensen Huang, Nvidia's creator and CEO, emphasized the importance of data in the AI revolution and expressed optimism that the relationship with Snowflake will allow organizations to optimize their proprietary data and realize its full potential through generative AI.

Snowflake, Nvidia, and Cortex AI

Snowflake and Nvidia collaborate to incorporate Nvidia AI Enterprise software capabilities into Cortex AI, allowing business customers to design and deploy customized AI-powered applications efficiently.

This integration comprises NeMo Retriever, which enables high-accuracy information retrieval for organizations within Cortex AI and Nvidia Triton Inference Server. This allows for simple deployment and scalability of AI inference.

Furthermore, Nvidia NIM inference microservices, part of Nvidia AI Enterprise, may now be created straight from Snowflake as a native app powered by Snowpark Container Services. This functionality streamlines the deployment of foundation models in Snowflake, making it easier for enterprises.