IBM Keeps a Sharp Storage Edge

IBM Edge is the company’s annual systems conference with a focus on servers (namely z Systems and POWER) and storage solutions. As you well know, such a conference is an immersive experience with the overwhelming amount of information that is presented within a few days. As an industry analyst with a strong focus on storage, I welcome the deluge, although sorting out what it all means can take some time. Still, here are some summary thoughts that I hope you will find of value.

Obviously, conferences like Edge stress new product announcements, but they also highlight the host company’s vision. Although some think that is unnecessary or impractical (note the George W. Bush “vision thing”), a company vision is both necessary and practical, as it commits the time of its people and its money for continued product development. A sense of balance is necessary between being the leading edge (where things are advanced and innovative, but are likely to be adopted over time) and the bleeding edge (which is often too far ahead of its time and likely to be too high risk for many vendors). As its 100 plus years have shown, IBM has basically been successful at the leading edge and Edge 2016 suggests this state of affairs continues.

Cognitive solutions and the cloud

Supporting and delivering IBM cognitive solutions on a cloud platform was one of the major themes and a leading vision of the conference. What does this entail? We all recognize the impact that the Internet of Things (IoT), social media, mobile applications and the like are having on modern businesses. These result in new workloads that require new software, usually considered to be in the nature of software-defined computing. This involves analytical tools as well as machine learning and deep learning capabilities that can deal with the vast amounts of data that come from a variety of sources at an increased velocity, which is often labeled as big data.

IBM has chosen to call this shift the Cognitive Era to reflect that it is much more than big data and to take advantage of its leadership strengths in cognitive technologies (such as the famous IBM Watson as well as its portfolio of advanced analytics products). IBM believes that the best way to deliver cognitive solutions is through hybrid cloud services offered through its own IBM/Softlayer Cloud platform.

Now where does IBM Storage play a role in this vision? An old term, data processing, means that computing (which is software-dependent) and storage (where the data resides) are intrinsically intertwined. IBM can enable the use of storage through three different deployment models — software-based, as a cloud service or as a physical system. The key is to deliver that data when, where and how the data is needed efficiently and at a reasonable cost, and  IBM used several customer examples to illustrate its point:

  • Coca-Cola improved performance for what IBM calls Cognitive Forecasting with 20x more forecasting data processed with a 97% reduction in processing time using a combination of IBM’s FlashSystem and Spectrum Virtualize.
  • MD Anderson, the largest cancer center in the United States, provides what IBM calls Cognitive Treatment by improving outcomes for cancer patients by matching case histories with best clinical trials using a combination of IBM Spectrum Scale and IBM Watson
  • SparkCognition provides what IBM calls Cognitive Security to improve response to cyber threats with a combination of flash (IBM FlashSystem) to accelerate cognitive applications (report processing reduced from 2 hours to 6 minutes) and AI (IBM Watson) to prevent cyber-attacks (resulting in a 72% reduction in threat identification query speed).

IBM’s storage focus

From a storage perspective, IBM continues to emphasize three things: 1) the breadth of its storage portfolio (which means that it can meet a wide range of customer use cases and workloads), 2) leadership in software-defined storage (SDS) where it can adapt to high performance or high capacity needs by leveraging the appropriate underlying storage media — flash, disk or tape — for file, block, or object storage as necessary, and 3) a leadership position in flash storage (where it views a swift transition for production applications not only for enhanced and more predictable performance, but also for operational efficiencies, including energy efficiency as well as the ability to eliminate the need for both storage and server overprovisioning).

IBM also tied these points to storing data on the hybrid cloud (a mix of on-premises and off-premises environments), the platform on which its cognitive solutions are built. Note that IBM does not neglect traditional application workloads, which can leverage automation through the use of flash and cloud to modernize. Instead, the company focuses on the transformation to new application workloads that are the workhorses of the Cognitive Era and can leverage software-defined storage in addition to flash technologies and IBM Cloud services.

Rather than go into detail on IBM’s storage portfolio, please see the following blog that is available at my Mesabi Group Web site.

However, a couple of newer products were frequently mentioned by IBM at Edge:

  • IBM Spectrum Copy Data Management — This is a new product in the Spectrum Storage SDS family; it is a stand-alone product and is complementary across all IBM Storage products; it has three principal areas of focus: 1) cloud, where workloads can either be migrated or replicated to the cloud, 2) data protection, for managing copies of data, including creating, tracking, refreshing, and deleting copies, and 3) DevOps, where, among other things, a copy of production data can be provisioned within minutes.
  • IBM Cloud Object Storage — Object storage is a major trend in storing fixed content cost efficiently at scale and is now a third member of a triad that already included block and file storage; IBM Cloud Object Storage is the name the company has given to the assets from its Cleversafe acquisition; Cleversafe was one of the acknowledged leaders in the object storage world; now IBM has extended the product to support object storage on-premises or in the cloud as a software, service or system offering, as appropriate.

Mesabi musings

Sometimes the question is asked of IBM of why it is still in the infrastructure business of servers and storage. My answer is that despite all the talk about the digital transformation, the heart of the matter is still “data processing” and that implies both computing and storage elements. Yes, the focus is (and rightfully so) on the importance of software, but the last time I looked, software does not run on the ether alone, but on physical systems that include both computing and storage elements.

While some servers and storage can be considered to be commodities, that would be far from the truth with IBM’s z Systems mainframes, Power Systems servers and storage portfolio, all of which have a broad set of characteristics that distinguish them from competitors’ offerings. From a storage perspective, even though the SDS products in IBM’s Spectrum Storage portfolio can work well with many heterogeneous storage hardware products, the company believes SDS works even better with its own products, say IBM FlashSystems.

So even as the Cognitive Era and the cloud continue to alter the IT universe, they also enhance, rather than distract from the need to have strong system architectures. That was the clear point and takeaway from IBM Edge 2016.