Oracle recently announced the availability of HeatWave GenAI, offering customers the ability to build generative AI applications without requiring AI expertise, data movement, or extra costs. The new features include the industry’s first database large language models, an automated database vector warehouse, scalable vector processing, and contextual natural language conversations. HeatWave GenAI enables developers to create a vector store for unstructured enterprise content using a single SQL command and built-in embedding models, allowing for natural language searches and contextual conversations within the database without needing GPUs. This innovation aims to reduce application complexity, enhance performance, improve data security, and lower costs.
The addition of HeatWave GenAI enhances Oracle’s existing integrated capabilities, providing developers an intuitive way to interact with their enterprise data and generate precise answers quickly. The new automated features, such as LLM in the database and automated vector warehouse creation, simplify the development of generative AI applications, improve search, retrieval, and content generation processes, and reduce complexity and costs for customers.
HeatWave GenAI integrates with other Oracle services, offering pre-trained foundational models from leading LLM providers and utilizing sematic queries with standard SQL. The scalable architecture supports very fast semantic search results, ensures accuracy, and enables parallel processing on multiple nodes for quick responses. The system also includes a visual code plugin, Heat Wave Chat, that facilitates natural language and SQL queries, file selection from object storage, and context maintenance for responses generated by LLMs.
Benchmark analyses confirm HeatWave GenAI’s superior performance compared to competitors like Snowflake, Databricks, and Google BigQuery. The service offers faster vector store creation with lower costs, efficient similarity search queries, and accurate results with predictable response times. Customers commend the democratization of generative AI through HeatWave GenAI, citing improved application productivity, reduced complexity, and enhanced customer experiences.
Collaboration between Oracle and AMD further enhances HeatWave GenAI’s performance by leveraging AMD EPYC processors. The integration of HeatWave’s automation, in-database features, and vector processing capabilities highlights Oracle’s innovation in cloud services and underscores its commitment to delivering advanced AI solutions for enterprise workloads.
Overall, HeatWave GenAI represents a significant step in Oracle’s goal of providing an all-encompassing cloud database solution that integrates generative AI, machine learning, and transactional analytics for businesses of all sizes. With its advanced features, cost-effectiveness, and performance benchmarks, HeatWave GenAI stands out as a powerful tool for developers looking to leverage generative AI in their applications.
Article Source
https://www.prnewswire.com/news-releases/oracle-announces-industry-first-in-database-llms-and-an-automated-in-database-vector-store-with-heatwave-genai-302182466.html