Jean is a senior industry analyst focusing her research on server technology, storage technology, database software and the emerging market for Software Defined Infrastructure (SDI).
IBM Systems’ year-end report to analysts was remarkable in the way that it positioned distributed computing, supporting hybrid cloud and analytics, as an important key to the company’s overall systems strategy. It is also vital for growing installations of IBM hardware and software, and expanding into new accounts.
IBM has long been involved in scale-up and scale-out computing, but its brand is often associated with traditional scale-up systems and storage for data centers.
This year-end report showed how the portfolio, as a whole, enables distributed computing and scale-out computing, to support hybrid cloud, analytics and cognitive computing. With this approach, IBM’s scale-out systems – seen in clusters and cloud services – coordinate with scale-up systems, such as IBM z Systems and scalable IBM storage. Overall, this positioning gives IBM ways to approach new customer sets adopting cloud that might not have considered IBM hardware and software before.
As cited by Tom Rosamilia, IBM senior vice president, Systems, three top examples of this portfolio approach for distributed computing are:
Hybrid Cloud Deployments
IBM is working to grow its customer set by tapping into end-to-end solutions that tie enterprise data centers – long associated with IBM’s scale-up systems – to cloud-delivered services featuring scale-out systems. These hybrid cloud solutions combine IBM systems platforms with IBM cloud object storage products and IBM Spectrum storage-management software. IBM has provided systems for hybrid clouds for years now, but is emphasizing hybrid cloud solutions based on new and enhanced hardware and software products and services.
It’s worth noting that, many years ago, IBM would have emphasized sales of its own x86 servers as platforms for these hybrid cloud solutions. Given IBM’s 2014 sale of its x86 server group to Lenovo, these cloud-based solutions give IBM another avenue to tap into cloud services running on x86 servers. They can do so through use of IBM’s SoftLayer IaaS services (for x86 and Power), or by using other cloud services provided by CSPs and MSPs. Importantly, they can use IBM distributed cloud-object software for hybrid cloud platforms, which stores and manages data on available resources, spanning multiple data centers.
Cognitive Computing and Analytics
Cognitive computing workloads with HPC (high performance computing) and analytics programs provide more ways for IBM’s Power systems business to grow beyond its traditional IBM/AIX Unix installed base.
IBM’s strong support for Linux, across IBM Systems platforms, will help drive this growth in cognitive computing workloads. Today, a substantial slice of IBM’s Power business is Linux-on-Power systems running HPC or SAP analytics. Likewise, LinuxONE systems support cognitive computing workloads. Rosamilia told analysts that Power deployments for SAP/HANA real-time analytics are spurring growth in new and existing IBM accounts.
Although the Systems call did not dwell on it, IBM’s market initiatives in Watson for cognitive computing and accelerated analytics can be other elements of an overall IBM cognitive solution for enterprise customers. Of course, IBM will not be alone in providing these types of solutions. We expect to see competition from Dell EMC, HPE and Oracle in these accelerated analytics workload opportunities.
There are many more technical details underlying the distributed computing strategies for cloud computing, hybrid cloud, analytics and cognitive that could not be described here, in a short blog item. There are also many distributed-computing partnerships with companies like Nvidia (for its GPU accelerators and high-speed NVLink); Xilinx (for co-processors) and Mellanox (for high-speed interconnects) – and the OpenCAPI interconnect consortium.
This blog takes the view that it is not the technology alone that gives momentum to this distributed computing initiative. Instead it is having an IBM Systems strategy centered on workloads, connectivity, and end-to-end solutions across infrastructure that is enabling a broader reach for IBM’s Systems point-products.
In short, IBM is no longer speaking of its systems platforms in isolation, pushing the technical specifications of each as the primary reasons for customers’ consideration of new Systems products. By supporting, and building on, distributed models for computing, IBM is showing the business value of its platforms in the broader world of end-to-end workloads spanning corporate and cloud data centers.
Last modified on