Enterprises have accelerated their adoption of AI over the past year. Gartner predicts that by 2026, more than 80% of enterprises will have AI APIs or AI-powered generative applications in production, up from less than 5% in 2023. However, an enterprise’s ability to get value from AI will depend on the availability and quality of the underlying data. To unlock the full value of data for AI, enterprises must be able to navigate complex IT environments to enable data access, optimize price-performance for workloads at scale, and prepare and deliver governed data for AI.
IBM® watsonx.data™ enables businesses to scale AI and analytics wherever their unique data resides. It is a core component of the IBM watsonx™ AI and data platform, which enables enterprises to create custom AI applications for specific business needs, access and manage all data sources, and implement responsible AI workflows. Allowing you to accelerate all on one platform.
Watsonx.data allows businesses to unlock the value of their existing data by connecting to their existing storage and analytics environments. You can also prepare data for your AI use cases and optimize the cost of your workloads with multiple, purpose-built query engines and low-cost object storage.
That's why we're excited to announce new and upcoming updates to IBM watsonx.data. Join us at Think 2024, an annual event that brings together over 5,000 technology pioneers and leaders. So what's new?
Achieving excellent cost performance
IBM watsonx.data and Presto C++ v0.286 and query optimizer on IBM Storage Fusion HCI has been internally tested by IBM and Better price performance Equivalent query execution times compared to Databrick's Photon engine less than 60% of the cost, Derived from the publicly available 100 TB TPC-DS query benchmark. *
Presto, an open source Linux Foundation project, is the primary engine for watsonx.data. Presto C++ Presto 2.0 is the next generation version of Presto being developed by Meta, IBM, and other companies to run Presto on Velox. Velox is an open source C++ native acceleration library designed to be configurable across compute engines. At IBM, he is the primary maintainer of the Velox project and has contributed to the development of Presto 2.0, including the Parquet reader, Iceberg reader, and file system support. query optimizer Integrates enterprise-proven query compilation technology with advanced query rewrite and cost-based optimization techniques. In other words, watsonx.data is enhanced for fast query time performance at optimized cost.
Unlock transactional mainframe data for AI and analytics
We announced the upcoming launch of IBM Data Gate for watsonx, a technology that will revolutionize how organizations synchronize, analyze, and build AI models from data generated on IBM Z®. A 2022 Celent report commissioned by IBM estimates that 70% of bank, card, and payment transactions worldwide occur in IBM zSystems™ environments.1 IBM customers will be able to unlock Transactional mainframe data for AI and analytics IBM Data Gate for watsonx™ is integrated with IBM watsonx.data.
Using this valuable transaction data, organizations can identify fraud, understand voter behavior, customer purchasing journeys, customer attrition, and build predictive AI models to understand and predict business outcomes. and move forward. By ingesting transactional data that originates on the mainframe into an open, managed data lakehouse like watsonx.data, businesses can easily build AI models that ultimately increase revenue, improve productivity, and Helps control costs.
Integrate and share data across IBM databases for new AI applications
From mobile banking applications to connected cars, customers rely on popular IBM databases to store their most important data across hybrid clouds and power the applications and analytics that run their businesses every day. Watsonx.data helps clients leverage this valuable data for AI. IBM DB2 database, Db2 Warehouseand infomix Coming soon is new integration with watsonx.data and support for open formats such as Apache Iceberg, which unifies and shares a single copy of data and metadata across hybrid clouds without the need for migration or re-cataloging. It's a schedule. Customers can also query data in IBM databases across multiple engines to prepare data for AI.
On-premises database clients can modernize to hybrid cloud deployments and get AI flexibility with SaaS-like compatibility. To support hybrid cloud application modernization, IBM and AWS also introduced consumption-based licensing. Amazon RDS for Db2on-demand licensing and faster cloud provisioning simplify workload management and accelerate time to market.
Accelerate data discovery and insights using a semantic layer – no SQL required
We've shared information about upcoming releases. semantic layer Can be embedded in IBM watsonx.data as part of IBM Knowledge Catalog. The semantic layer uses large-scale language models (LLMs) to create a unified data context across IBM data and AI tools. A semantic layer powered by watsonx not only enriches your data, but also provides automation tools to help your team quickly explore and process data.
When embedded in IBM watsonx.data, the semantic layer can generate data enrichments that enable clients to search and understand previously unsolvable structured data in natural language through semantic search, accelerating data discovery and unlocking data insights faster – no SQL required.
Scale repeatable data packaging and delivery for AI use cases
Announced IBM Data Product Hub is a new data sharing solution scheduled to be available in June 2024 that helps businesses achieve data-driven outcomes by streamlining data sharing between internal data producers and data consumers to access data. We will help you accelerate. What does this mean for IBM watsonx.data users? Connect to IBM watsonx.data to gain integrated access to disparate data sources, ingest relevant metadata, and create the core of reproducible, managed data products. Its data products enable you to deliver the right data for a variety of AI use cases at scale across your organization without having to run repetitive manual workflows and slow things down.
Create a knowledge base to make AI more relevant and accurate
Using trusted, controlled data is essential to ensuring the accuracy and relevance of AI applications.Therefore, we recently integrated vector database Based on open source Milvus In IBM watsonx.data. WatsonX clients can now integrate, curate, and prepare vectorized embeddings for generative AI applications at scale across trusted, managed data. This improves the relevance and accuracy of AI outputs such as chatbots, personalized recommendation systems, and image similarity search applications. This allows clients to seamlessly connect to trusted data in watsonx.data from IBM watsonx.ai™ or another AI tool.
How can I start using IBM watsonx.data now?
Try watsonx.data with a free trial. Want to learn more about upcoming product updates? Join us for upcoming webinars featuring Think announcements or schedule a meeting with an IBM watsonx.data product specialist.
Explore Watsonks now
* Based on IBM internal testing of Presto C++ 0.286 in a hyperconverged infrastructure setup with 1 master + 75 worker nodes, 1009 vCPUs, 18 TB memory, 344.8 TB filesystem storage, distributed RAID, and 50 GB network, public Databricks Compared to 100 TB TPC – DS Query benchmark published in 2021 with 1 master + 256 worker nodes, 2112 vCPUs, 16.1 TB memory, 528.2 TB total storage, 10 GB networking. Pricing calculations are based on IBM watsonx.data pricing as of May 7, 2024 and Photon pricing published by Databricks as of May 7, 2024. Results are based on testing conditions and prices as of the date indicated. Actual costs and performance may vary depending on individual client configuration and conditions. The results are from the Databricks SQL 8.3 benchmark, and the results are not compliant with the Databricks SQL 8.3 benchmark specification and cannot be compared to publicly available Databricks SQL 8.3 benchmark results.
1 Celent Report: “Operating Fraud Prevention on the IBM Z16, Neil Katcoff,” April 5, 2022, Commissioned by IBM
