Oracle is going after the lakehouse market with HeatWave to go along with a bevy of generative AI features including HeatWave GenAI and HeatWave on AWS. Oracle also launched its Intelligent Data Lake and genAI apps across its platform. Oracle also announced a zettascale cloud computing cluster with Nvidia's Blackwell platform.
At Oracle CloudWorld in Las Vegas, Oracle introduced HeatWave Lakehouse, HeatWave GenAI, HeatWave on AWS and HeatWave MySQL enhancements. MySQL became part of Oracle via the Sun Microsystems acquisition in 2009. In 2020, Oracle launched HeatWave a cloud-native in-memory query accelerator designed to speed online analytical processing (OLAP) to deliver real-time analytics and other complex queries within the MySQL database as a managed service in Oracle Cloud Infrastructure (OCI).
The HeatWave news landed a day after Oracle announced a deal with Amazon Web Services (AWS) for Oracle Database@AWS. With the move, Oracle has database partnerships with all of the hyperscale cloud providers.
Here's a look at the HeatWave announcements:
HeatWave on AWS. Oracle said it is launching HeatWave GenAI and HeatWave Lakehouse on AWS as well as OCI. The core pitch is that Oracle HeatWave can deliver better performance. Oracle said users can automate vector store creation and vector embedding, use large language models in-database running on CPUs as well as Amazon Bedrock and have natural language conversations with documents in Amazon S3.
According to Oracle, HeatWave vector processing offers better price performance than Snowflake, Databricks and Google BigQuery.
Research: Oracle MySQL HeatWave Grows, Adds Lakehouse Support
HeatWave Lakehouse will give AWS insights on structured, semi-structured and unstructured data on Amazon S3 and offer native JavaScript support and give AWS users the ability to predict the right set of indexes for their OLTP workloads.The AWS moves round out Oracle's multicloud strategy.
- Oracle partners with Google Cloud, takes on OpenAI workloads
- Oracle's MySQL HeatWave additions take aim at AWS with lakehouse support
- Oracle adds generative AI features into Fusion Cloud, expands Azure partnership into Europe
- Palantir will move workloads to Oracle Cloud as both court governments and enterprises
HeatWave GenAI will give customers multi-lingual support to load documents in 27 languages into HeatWave Vector Store. Oracle is also adding optical character recognition (OCR) support to HeatWave Vector Store, LLM inference batch processing and automatic vector store updates.
HeatWave Lakehouse will give customers the ability to query data in object storage at the same speed as database queries. Users can write results to object storage and use HeatWave for MapReduce applications. Oracle said it is also adding automatic change propagation to HeatWave Lakehouse.
HeatWave MySQL will get the ability to optimize query plans and improve performance, integration with OCI Ops Insights and bulk ingest.
HeatWave AutoML will be able to build, train and explain machine learning models in HeatWave without additional costs. Oracle said HeatWave AutoML is getting the ability to store and process larger models, model topics, manage data drift and add semi-supervised log anomaly detection.
All OCI accounts get access to a standalone HeatWave instance in OCI home regions with 50 GB of storage and 50 GB of backup storage for an unlimited time.
Constellation Research's take on HeatWave additions
Constellation Research analyst Holger Mueller said it was critical that Oracle added genAI and lakehouse functionality to HeatWave.
"Oracle has to make it easy enough for Oracle DB customers to stay with their Oracle databases. If Oracle succeeds it will keep customers using its databases. If Oracle made it hard, customers would look for database and lakehouse alternatives – which would not be a good outcome for Oracle. Lakehouse is critical--Oracle has been absent from it so it's very much needed. The multicloud deployments of HeatWave are as well."
"For HeatWave GenAI, the update is significant. Oracle had to add vector support, and this release is all about to make it easier for developers to use vector capabilities inside HeatWave. Oracle added JavaScript support which is a big step. Basically, Oracle needs to make sure that the data content in HeatWave is available and it is easy for developers to use the vector support. If the latter succeeds the future of HeatWave in the AI era is set."
Constellation Research analyst Doug Henschen said:
"Oracle's latest announcements on MySQL Heatwave step up the competition in the data warehouse/data lakehouse market, particularly with AWS. While AWS continues to focus Aurora on transactional needs, Redshift on analytical needs and SageMaker on data science, the combination of HeatWave on AWS, HeatWave Lakehouse, HeatWave AutoML, and HeatWave GenAI brings together a compelling set of capabilities on a single platform."
Oracle Supercluster on Nvidia Blackwell
Oracle announced a zettascale cloud computing cluster with Nvidia's Blackwell platform. OCI said it is taking orders for the AI supercomputer, which has up to 131,072 Nvidia Blackwell GPUs available.
According to Oracle, the AI supercluster has 2.4 zettaFLOPS of peak performance and can outperform the Frontier supercomputer. Oracle said:
"OCI Superclusters are orderable with OCI Compute powered by either NVIDIA H100 or H200 Tensor Core GPUs or NVIDIA Blackwell GPUs. OCI Superclusters with H100 GPUs can scale up to 16,384 GPUs with up to 65 ExaFLOPS of performance and 13Pb/s of aggregated network throughput. OCI Superclusters with H200 GPUs will scale to 65,536 GPUs with up to 260 ExaFLOPS of performance and 52Pb/s of aggregated network throughput and will be available later this year. OCI Superclusters with NVIDIA GB200 NVL72 liquid-cooled bare-metal instances will use NVLink and NVLink Switch to enable up to 72 Blackwell GPUs to communicate with each other at an aggregate bandwidth of 129.6 TB/s in a single NVLink domain."
Nvidia Blackwell GPUs on the supercluster are available in the first half of 2025.
Oracle Intelligent Data Lake, GenAI analytics on Oracle Data Intelligence Platform
Oracle launched Oracle Intelligent Data Lake to go with its Oracle Data Intelligence Platform.
As Mueller noted, Oracle's plan is to surround its databases with all of the key components that would attract existing customers to other platforms.
Oracle said that Intelligent Data Lake will be a core component to the Data Intelligence Platform. Oracle is looking to provide a unified experience by combining orchestration, data warehouses, analytics and AI within the Data Intelligence Platform, which runs on OCI.
The Data Intelligence Platform will include integration with Oracle Autonomous Data Warehouse, Oracle Analytics Cloud, HeatWave, AI services and third-party services.
According to Oracle, the Intelligent Data Lake will enter "limited availability in 2025."
In addition, Oracle is embracing open standards with its Intelligent Data Lake, which will support data catalogs, Apache Spark, Apache Flink, Jupyter Notebooks and open-source standards such as Kafka, Delta Lake, Iceberg and Parquet.
Oracle also said its Oracle Analytics Cloud AI Assistant is available and Autonomous Database will support retrieval augmented generation (RAG) and new no code tools as part of Data Studio.
Oracle Fusion Data Intelligence, HCM, SCM apps
Oracle Fusion Data Intelligence has intelligent applications for Oracle Fusion Cloud Human Capital Management (HCM) and Oracle Fusion Cloud Supply Chain & Manufacturing (SCM) designed to recommend actions. Fusion Data Intelligence combines data, analytics, prebuild AI and machine learning models to create actionable insights to enterprises.
The company added the following:
- Oracle Cloud HCM gets a People Leader Workbench app that aims to align business and financial goals for HR and finance execs.
- Oracle Cloud SCM gets Supply Chain Command Center to help enterprises respond to changing conditions across the supply chain network.
- Fusion Data Intelligence gets operational reporting as well as an AI-powered developer assistant.
- ERP, HCM, SCM and CX analytics get new AI and machine learning feature.
GenAI Agents with RAG
Oracle announced the general availability of OCI GenAI Agents with RAG capabilities. OCI GenAI Agents access Oracle Database 23ai Vector Search and add an automation layer.
The company said the use cases for OCI GenAI Agents include call center optimization, legal research, revenue operations and HR recruiting functions.
In addition, OCI GenAI will be able to access Meta Llama 3.1 models as well as Cohere's Command R, Command R+ and Embed models. OCI Data Science will be able to tap into OCI Ampere A1 as well as models from Hugging Face.
Oracle also added document understanding, support for more than 100 languages, vision and speech capabilities as well as Code Assist, which is in beta.
Generative development
Oracle launched generative development (GenDev) that combines Oracle 23ai and multiple Oracle services and support for LLMs.
The company also launched Autonomous Database Nvidia GPU support with GPU-enabled Python packages, Data Studio AI enhancements, Autonomous Database for Developers and a version with container images.
Oracle also announced Autonomous Database Select AI, which supports synthetic data in test instances.
Cloud optimization
Oracle launched Cloud Success Navigator, a platform that optimizes cloud and AI deployments. Cloud Success Navigator, which is available at no cost for 12 months, includes:
- Preconfigured environments for specific enterprise processes.
- Best practices from Oracle's library of business processes, process flows, learning content and guides.
- Deployment guidance by roles.
- Recommended actions for cloud quality.
- Dashboards for implementation status and milestones.
Oracle Fusion Cloud Applications
Oracle outlined a series of AI tools across finance, supply chain, HR, sales, marketing and service functions in the Oracle Fusion Cloud Applications Suite. Here's the breakdown:
- Oracle Cloud ERP gets predictive cash forecasting tools via AI models that operate across multiple time frames s well as narrative reporting toos, variance explanations and commentary. Transaction records will be automated.
- Oracle Cloud HCM as a set of bespoke AI skills that combines skills data with enterprise and third-party data.
- Oracle Cloud SCM adds a new smart operations workbench and assisted authoring in Oracle Order Management.
- Oracle Cloud CX has generative AI tools to answer contract questions, write emails and summarize for quotes and proposals.
Oracle Unity Customer Data Platform (CDP), part of Oracle Fusion Cloud CX, will get an account profile explorer to spot revenue opportunities, buying group and opportunity scoring, native Oracle Analytics Cloud integration and industry optimized templates, data models and attributes.
