IBM ESS for AI and Big Data storage solution
Use Cases and Deployment Scope
IBM Elastic Storage is a highly available and highly scalable software based storage system for AI and Big Data. We had high performance use cases for our AI applications for an AI based product that we were building and IBM Elastic storage looked like a perfect solution.This service simplified the storage for Data and Model management and we were easily able to storage huge multiple types of data include images, natural language as well as structured datasets using this solution.
Pros
- Very High Availability and Scalability
- Works particularly well for AI and Big Data usecases
- Easily unifies all sorts and types of Data
Cons
- Integration with other Big Data Platforms like EMR and Databricks needed
- May not be suitable for non AI related/Legacy usecases
- Native support for structured relational datasets
Most Important Features
- High scalability for AI and Big Data usecase
- Integration with IBM Data platform
Return on Investment
- Faster read/write data pipeline workloads
- Faster time to load models from Storage
- Improved inference time for models
Alternatives Considered
Amazon S3 (Simple Storage Service), Databricks Lakehouse Platform (Unified Analytics Platform) and Amazon EMR (Elastic MapReduce)
Other Software Used
Amazon S3 (Simple Storage Service), Databricks Lakehouse Platform (Unified Analytics Platform)
