Abstract
The IBM z14 is designed for cognitive and analytics processing to support business use cases. With the explosive growth in the amount and richness of data and data-driven business processes, being able to process, analyze, and apply machine learning in a timely and efficient manner is often a business necessity. Improvements to the single-instruction-multiple-data (SIMD) facility, as well as improvements to Java garbage collection, provide improved support for different software packages. Software such as Apache Spark has been enabled for both z/OS and Linux on z to provide the software infrastructure required for cognitive and analytics processing, while new software such as IBM Machine Learning for z/OS provides additional assistance for data scientists. End-to-end cross-platform solutions that require data that is stored on the z14 can also be improved using these new software and hardware capabilities. This paper explores the various hardware capabilities of z14, as well as the enabling software and end-to-end solutions.