IBM Power

When it comes to IT, companies need to get as much return on investment as possible. IT infrastructures must be sufficiently adaptable to react flexibly to changing processing needs and constantly increasing workloads, without leading to further complexity. But flexibility and performance alone are not enough. In addition, IT infrastructures need to offer a high degree of security and provide virtually uninterrupted availability. At the same time, costs for energy and cooling need to be reduced. IBM Power Servers offer all this: top-class performance, flexibility, reliability and cost savings.

Technology and performance

Through the continuous development of new technologies for IBM Power Systems, IBM has a reliable roadmap and offers you protection for your investments going forward. You can quickly amortize your IT investments while also benefiting from services and cost benefits over the long term.

The processors of current IBM POWER9 systems offer much greater memory bandwidth per socket than x86-based systems. This is a huge advantage when using an in-memory database like SAP HANA.

The extraordinary performance of the POWER9 architecture allows businesses to process transactions and queries with constantly low loading times, even under very high load. A higher throughput per core enables businesses to meet the performance requirements for SAP HANA, for example, or a significantly higher Linux container workload, with fewer processor cores.

But even the I/O performance, critical in log writing, data persistence and starting databases (e.g. SAP HANA, D2b or Oracle), for example, is much greater on IBM Power Systems than Intel-based ones.

Flexibility

You can operate mixed workloads together in a shared processor pool. The workloads share the available processor cycles, enabling flexible usage and automated resource distribution according to requirements, without the need for manual interventions. In addition, users can freely select the size of the VMs (RAM, core) and change them quickly online at any time. It is also possible to expand or reduce the required I/O components, in order to adjust these to your workloads.

Thanks to PowerVM Live Partition Mobility (LPM), you can clear entire systems online for planned maintenance work or combine workloads like an SAP application server on the same server as the database instance. This improves the latencies across the local virtual network. LPM can also be used for the online migration across different POWER processor architectures, in order to achieve an interruption-free transfer to a new processor generation (e.g. from POWER8 to POWER9). This is one of the Power platform’s true unique selling points on the market.

Reliability and availability

The globalization of markets across various time zones means that there is often no longer a window available for maintenance work. If there is, it is frequently so small that the work can no longer be carried out in full. IBM Power Servers allow most maintenance activities to be performed during ongoing operations. This drastically reduces downtimes for planned maintenance and minimizes service losses.

Power Servers offer high reliability for business-critical applications and are based on the best RAS architecture on the market (e.g. unique architecture with mirrored hypervisor for querying memory faults). This means that Power Servers have the highest availabilities alongside those of the IBM Mainframe, at a percentage value of 99.999. As a result, you can run your business-critical applications around the clock even more reliably.

Cost reduction

Consolidating underutilized servers on highly efficient IBM Power Servers grants massive potential for savings. You can therefore benefit from higher utilization and resource sharing to support your applications. In addition, IBM’s capacity-on-demand concept makes it possible to clear resources (processor and RAM) on a Power Server and use them for load peaks. Users can also share the resources in a server pool for even greater flexibility. Consumption is measured on a minute-by-minute basis, while costs can be shifted from CAPEX to OPEX.