June 19, 2020 By Kevin Oliver 3 min read

As companies progress on their Journey to AI, there is considerable focus on what needs to be available to build AI driven applications. The rungs of the AI ladder, which are best described as Collect, Organize, Analyze, Infuse, and Modernize are designed to strengthen a company’s use of AI. However, the individual impact of AI on each rung should not be overlooked. AI can be both the goal and the means of achieving that goal

Take the collect rung, which represents hybrid data management, as an example. A database should be built for AI with support for popular language use, natural language querying, graph, and blockchain. But it should also be powered by AI to optimize data management for greater efficiency and performance. The eBook, “Db2 – The AI Database” covers eight technologies that make a database built for AI and powered by AI. Two “powered by AI” technologies, Adaptive workload management and resource optimization as well as machine learning query optimization are covered in greater detail below.

Adaptive workload management and resource optimization

Adaptive workload management was created to more intelligently allocate data resources across workloads either automatically or by providing an alert based on the business’s preference. Machine learning is an integral part of this process, monitoring expected and actual runtimes so that predictions for future workloads can be made and resources can be adjusted accordingly. If it is likely that an issue will occur, adaptive workload management can notify users of the finding or automatically take steps to correct the problem.

This is in contrast to previous techniques for allocating resource allocation where a pre-determined limit for workload number or size was set by users. These limits would then need to be monitored and regularly adjusted manually in an attempt to keep at top efficiency. IBM testing has found the better utilization possible with adaptive workload management to result in database performance improvements up to 30 percent.

This is partially due to the fact that it takes the previously highly manual and labor-intensive process of tuning and helps automate it. DBAs can use more of their time on value-additive activities rather than maintenance. It also helps avoid under-utilization of assets and performance issues or failures that may occur due to human error.

Machine learning query optimization

Additional optimization can be gained by applying machine learning to how queries are run. Traditionally, cost-based optimizers have been used. They worked well to recommend the fastest execution strategy in some cases, but weren’t great at understanding recent database changes and had no way to learn from experience. It would suggest the same strategy even if it hadn’t worked optimally on the previous occasion.

Machine learning query optimization changes this by learning from actual query performance and iterating on the suggestion it makes for which path the query should take. In this way, it mimics neural network patterns to learn from experience.

For example, a cost optimizer looking at four tables may suggest joining two pairs of tables together and then joining the results of those joins together. Machine learning query optimization could learn that this isn’t optimal and suggest joining two tables, then a third to the result, followed by a fourth to the result of that. IBM Internal testing has found that machine learning query optimization has resulted in some queries being completed 8-10 times faster.

In addition, as with adaptive workload management, the time that would have been spent by a DBA to monitor performance and make corrections can be put towards things that provide additional value.

Acquiring the AI needed to optimize data management

Companies looking to infuse AI throughout their organization, or those simply looking to optimize their information architecture, need to start with data management. IBM Db2 is a database built for AI, with capabilities powered by AI, that enable businesses to do both.

Adaptive workload management and machine learning query optimization are both core components of IBM’s data management strategy anchored by IBM Db2, which is available on premises, in the cloud and on IBM Cloud Pak for Data, which is built on Red Hat OpenShift Container Platform.

To learn more about technologies IBM uses to deliver data management that’s both powered by AI and built for AI read our latest eBook, “Db2 – The AI Database.” It has more information on adaptive workload management, machine learning query optimization, and six additional features positioned to help you succeed on the Journey to AI.

Read the AI Database eBook

Got Questions? Ask our Experts!

Schedule a free one-on-one consultation with our experienced data professionals and distinguished engineers who have helped thousands of clients build winning data management strategies.

More from Artificial intelligence

Best practices for augmenting human intelligence with AI

2 min read - Artificial Intelligence (AI) should be designed to include and balance human oversight, agency, and accountability over decisions across the AI lifecycle. IBM’s first Principle for Trust and Transparency states that the purpose of AI is to augment human intelligence. Augmented human intelligence means that the use of AI enhances human intelligence, rather than operating independently of, or replacing it. All of this implies that AI systems are not to be treated as human beings, but rather viewed as support mechanisms…

IBM watsonx AI and data platform, security solutions and consulting services for generative AI to be showcased at AWS re:Invent

3 min read - According to a Gartner® report, “By 2026, more than 80% of enterprises will have used generative AI APIs or models, and/or deployed GenAI-enabled applications in production environments, up from less than 5% in 2023.”* However, to be successful they need the flexibility to run it on their existing cloud environments. That’s why we continue expanding the IBM and AWS collaboration, providing clients flexibility to build and govern their AI projects using the watsonx AI and data platform with AI assistants…

Watsonx: A game changer for embedding generative AI into commercial solutions

4 min read - IBM watsonx is changing the game for enterprises of all shapes and sizes, making it easy for them to embed generative AI into their operations. This week, the CEO of WellnessWits, an IBM Business Partner, announced they embedded watsonx in their app to help patients ask questions about chronic disease and more easily schedule appointments with physicians. Watsonx comprises of three components that empower businesses to customize their AI solutions: watsonx.ai offers intuitive tooling for powerful foundation models; watsonx.data enables…

Announcing watsonx.ai & SingleStore for generative AI applications

2 min read - In December 2021, IBM and SingleStore announced their strategic partnership with SingleStoreDB with IBM and in December 2022 the partnership launched SingleStoreDB as a Service with IBM available on AWS, Azure and the Microsoft Azure Marketplace.  Now they are taking the next step in their strategic partnership to announce SingleStoreDB’s powerful vector database functionality’s support of watsonx.ai. Why watsonx.ai? IBM watsonx.ai is the next-generation enterprise studio for AI builders. It brings together traditional machine learning and new generative AI capabilities…

IBM Newsletters

Get our newsletters and topic updates that deliver the latest thought leadership and insights on emerging trends.
Subscribe now More newsletters