IBM Power Systems

IBM Power Systems are a family of server computers developed by IBM, integrating POWER processors, advanced virtualization, and enterprise-class reliability. They serve industries requiring high-performance computing, mission-critical databases, and large-scale virtualization, often deployed alongside platforms from Oracle Corporation, Hewlett Packard Enterprise, Dell Technologies, Cisco Systems, and Lenovo. Enterprises commonly pair them with software from Microsoft, Red Hat, SUSE, SAP SE, and Oracle Corporation for transactional, analytical, and cloud workloads.

History

The lineage traces to the POWER microprocessor project at IBM Research and corporate units such as IBM Rochester and IBM Hursley, continuing through mergers and collaborations with companies like AIX's originators and partnerships involving Lotus Development Corporation and Rational Software. Major milestones include consolidation of IBM's server lines after acquisitions of RISC System/6000 technology and subsequent integration with POWER-based systems used by customers including NASA, CERN, Bank of America, Walmart, and JPMorgan Chase. The 2008 announcement merged IBM's System p and System i product lines into the current family, following strategic decisions influenced by industry shifts involving Sun Microsystems, Compaq, SGI, and NEC Corporation. Over time, IBM collaborated with ecosystem partners such as Broadcom Inc., NVIDIA, Marvell Technology Group, Synopsys, and Cadence Design Systems to advance chip design, packaging, and interconnect technologies.

Architecture and Hardware

Power Systems are built around IBM's POWER architecture developed at IBM Research and manufactured with foundry partnerships including GlobalFoundries, TSMC, and Samsung Electronics. The architecture emphasizes multi-core, multi-threaded designs, NUMA topologies, and coherent cache hierarchies used in deployments by institutions like Argonne National Laboratory and Lawrence Livermore National Laboratory. Hardware innovations include chiplet designs and high-bandwidth interconnects such as technologies from OpenCAPI, CAPI Consortium, PCI Express, and interfaces influenced by InfiniBand Trade Association. Systems integrate accelerators and GPUs from NVIDIA and FPGA offerings often sourced in collaborations with Xilinx (now AMD), and employ storage controllers and SAN fabrics interoperable with products from EMC Corporation, NetApp, Pure Storage, and Hewlett Packard Enterprise. Reliability features trace concepts from earlier fault-tolerant efforts exemplified by projects at Bell Labs and Hewlett Packard research groups, while cooling and power designs reflect advances from Schlumberger and hyperscale operators like Amazon.com and Google.

Operating Systems and Software

Power Systems run operating systems optimized for POWER architecture, including AIX (operating system), IBM i, and Linux distributions supported by vendors such as Red Hat, SUSE, and Canonical (company). Enterprise middleware and database software commonly used on Power Systems include IBM Db2, Oracle Database, SAP HANA, Microsoft SQL Server (via virtualization), and analytics platforms from SAS Institute and Teradata. Virtualization and cloud stacks leverage technologies from VMware, Inc., KVM, PowerVM, and orchestration platforms like Kubernetes and OpenStack in collaborations involving Cloud Foundry and Red Hat OpenShift. Development toolchains and compilers originate from organizations such as GCC, LLVM Project, IBM Rational, and Eclipse Foundation, with performance tuning informed by workloads benchmarked by SPEC.org and applied research from NIST.

Performance, Scalability, and Use Cases

Power Systems target high-throughput transaction processing, analytics, and technical computing use cases adopted by Bank of America, Barclays, Morgan Stanley, Citigroup, and Goldman Sachs. They compete on metrics reported to organizations like SPEC CPU, Transaction Processing Performance Council, and are selected for supercomputing centers such as Oak Ridge National Laboratory and Argonne National Laboratory. Use cases include enterprise resource planning on SAP SE stacks, large-scale databases for Facebook, Twitter, and content delivery services used by Netflix, as well as scientific simulations for projects at Lawrence Berkeley National Laboratory and climate modeling groups at NOAA. Scalability strategies mirror practices from distributed systems research at MIT, Stanford University, Carnegie Mellon University, and University of California, Berkeley.

Models and Product Line Evolution

Product evolution spans generations with model families succeeding one another similar to transitions seen at Intel Corporation and AMD, influenced by ecosystem dynamics involving ARM Holdings and standards groups like the OpenPOWER Foundation. Notable platform updates introduced energy-efficient designs, integration of NVMe storage, and GPU-accelerated variants aligned with offerings from NVIDIA DGX and converged systems by Dell EMC. Commercial adoption patterns reflect migrations observed with platforms from Oracle Corporation after its acquisition of Sun Microsystems, and collaborations within consortia that include Google, Facebook, and Microsoft to shape open hardware initiatives. Current portfolios continue to be positioned for hybrid cloud scenarios alongside services from IBM Cloud, Amazon Web Services, Microsoft Azure, and Google Cloud Platform.

Category:IBM hardware