IBM Spectrum Computing – Brand New Brand for SDI

By Charles King, Pund-IT, Inc.  June 8, 2016

Product branding and rebranding may seem simple to outsiders, but for people inside the process it’s serious in the extreme. That’s partly due to the stakes involved—the number of folks who have lost their jobs and reputations in failed branding efforts are legion. But it’s also due to the more or less ineffable nature of the branding process.

In the IT industry, brand recognition tends to be a cumulative process that occurs over months and years of client relationship building. Customers like products they get to know and that provide satisfaction over time. But there are also situations where fundamental market or technological shifts make it imperative to alter a brand, reflecting how the vendor is adapting to those changes.

IBM’s recent introduction of its new Spectrum Computing platform and solutions followed the February launch of its Spectrum Storage Suite, and both aim to better define the company’s position in software-defined infrastructure (SDI) solutions. Let’s consider why that’s important for IBM and its customers, and how well “Spectrum” will work as a brand for SDI.

The SDI journey

“Software-defined” solutions for compute, storage and networking have such a high profile in contemporary IT markets that it’s easy to forget their foundational technologies have been around for decades. In short, SDI solutions use sophisticated software to abstract, manage and utilize complex hardware assets with the goal of gaining greater efficiencies and cost savings.

For example, many software-defined computing solutions depend heavily on the hypervisors underpinning various virtualization platforms, as well as tools for resource scheduling, workload management and other core functions. These technologies were originally introduced in the 1980s for mainframe systems, then adapted to other scale-up platforms over time.

In 1999, start-up VMware introduced virtualization solutions for Intel-based workstations, then extended those offerings to scale-out x86 servers in 2001. Over time, additional resource and workload management technologies were developed or adapted for x86 systems and supported new generations of cluster, grid and high performance computing (HPC) use cases.

Those developments, with the addition of sophisticated automation tools, enabled Google, Amazon and other innovators to extend software-defined management and control across ever larger scale-out compute environments. That included entire data centers that provide the foundation for an ever-growing variety of public cloud services.

How software-defined technologies help businesses continually adapt to ongoing changes is the point of this history lesson. So software definition should be regarded as an ever-evolving journey rather than an unchanging, ultimate destination. As computing technologies adapt to and are adopted for a growing range of use cases and processes, the tools for best-managing hardware assets must adapt, too. That’s the backstory in SDI.

Putting color in Spectrum Computing

So what does any of this have to do with IBM’s Spectrum Computing portfolio? Several things, but first and foremost is how the company’s new, updated SDI offerings aim to help IBM customers adopt new generation cloud-native applications and open source frameworks.

For example,

  • IBM Spectrum Conductor blends the intelligent resource and workload management features of Spectrum Computing with the data management and protection features of Spectrum Storage. Customers can use the platform to share compute and data resources to support complex applications. Plus, it enables administrators to leverage multi-tier hybrid storage strategies, including placing/accessing data on public cloud platforms, including IBM Cloud (Softlayer), Cleversafe object storage and Amazon S3.
  • IBM Spectrum Conductor with Spark is a specialized version of the platform that is designed to simplify the adoption and use of the popular open source Apache Spark framework for in-memory analytics. Simplification is its primary benefit but, according to IBM, the new solution can also speed analytics results by up to 60 percent.
  • IBM Spectrum LSF is a new batch/workload scheduler that works across distributed, heterogeneous compute infrastructures and HPC clusters. Features, including policy driven license management, simplified application integration and a comprehensive operational dashboard are designed to increase productivity and administrative efficiency.

While IBM Spectrum Computing solutions are designed to support scale-out, x86-based systems and infrastructures they can also be utilized in many IBM system environments. For example, IBM Spectrum Conductor fully supports IBM Power Systems solutions, including those recommended for Apache Spark. IBM Spectrum LSF is supported on both Linux on Power and IBM z System platforms, and is also aligned with IBM Power and OpenPOWER around HPC use cases.

Final analysis

So what does branding/rebranding have to do with IBM’s Spectrum Computing? Just as Spectrum Storage leverages some previously available company technologies, including the General Parallel File System (GPFS, now called Spectrum Scale) and tools from its XIV and Tivoli families, Spectrum Computing makes broad use of solutions and tools from IBM’s Platform Computing offerings. Those, in turn, were derived in large part from its 2011 acquisition of Platform Computing, a leading light in cluster, grid and cloud management software for distributed computing environments.

That does not mean that the Spectrum portfolio is a simplistic renaming exercise—anything but. IBM doesn’t manage acquired companies statically but instead provides those organizations and workers the assets they need to continue evolving and improving. As a result of those ongoing efforts, IBM has a host of popular, fully mature and supported technologies it can effectively leverage in SDI scenarios.

Moreover, the company has made significant investments in analytics technologies for a decade. That and its enthusiastic support for numerous open source projects and communities, including Apache Spark and Mesos, means that IBM Spectrum Computing is ready, willing and waiting at Ground Zero to assist businesses hoping to capture the advantages these new technologies promise.

That doesn’t mean that IBM Spectrum Computing or IBM Spectrum Storage will be automatic successes. Branding and rebranding efforts succeed to the extent that the involved vendors and their partners effectively engage customers and educate them about new products and services. That said, the past successes the company has enjoyed with products and technologies now in the Spectrum portfolio bodes well for the future.

The term “spectrum” is often used to describe a full range of colors—an artist’s palette rather than single, differentiated hues. In that same sense, IBM’s Spectrum portfolio contains a full range of solutions that can be used in a host of software-defined applications and use cases today. Plus, as the SDI landscape and organizations’ requirements evolve, IBM will evolve and adapt its Spectrum solutions to support those changes.

© 2016 Pund-IT, Inc. All rights reserved.