IBM has reportedly announced the launch of IBM POWER10, IBM’s next generation of IBM POWER CPU family. The new processor has been developed for providing a platform to fulfil the unique requirements of business hybrid cloud computing, and effectively deploys a design that is focused on performance and energy efficiency in a 7nm form factor, having an expected improvement of up to 3 times greater container density, processor energy efficiency, and workload capacity over the company’s IBM POWER9 processor.
According to Stephen Leonard, the General Manager at IBM Cognitive Systems, enterprise-grade hybrid cloud needs a robust off-site and on-premise architecture that includes co-optimized software and hardware. He has further stated that the company has developed IBM POWER10 as the premier processor that offers enterprise hybrid cloud, and effectively delivers the security and performance that customers expect from IBM. With the company’s stated objective of making Red Hat OpenShift the default alternative for hybrid cloud technology, the IBM POWER10 processor brings security enhancements and hardware-based capability for containers to the IT infrastructure level, added Leonard.
The IBM POWER10 processor has been designed over a period of five years with hundreds of new and pending patents and is considered as an important evolution in the company’s roadmap for POWER. Sources state that the systems built using the IBM POWER10 processor will be available during the second half of 2021.
According to reports, IBM POWER10 has been equipped with new hardware-enabled security capacities comprising transparent memory encryption that supports end-to-end security. Further, the processor has been effectively engineered to achieve remarkably faster encryption performance with 4 times the number of AES encryption engines per core in comparison to the company’s IBM POWER9.
IBM POWER10 processor has been further designed to support multi-petabyte memory clusters via Memory Inception, a breakthrough technology focused on the improvement of cloud capacity and economics for memory-intensive workloads from ISVs comprising the SAS Institute, SAP, and others in addition to large-model AI inference.