By Joe Clabby, Clabby Analytics
There are momentous changes taking place at IBM – changes that the press, some research analysts and many customers have yet to comprehend. At IBM’s InterConnect conference in Las Vegas last week, some of these changes began to surface. They include a major shift to cognitively-enabling the IBM application portfolio, as well as three major cloud announcements.
In opening keynotes, IBM started to allude to the importance of cognitive computing as a key component in future IBM product designs. The company made it clear that it is now starting to “cognitively-enable” several software offerings in its portfolio.
From a cloud perspective, IBM announced that: 1) it has cloud-enabled all of its major software offerings and now offers much of its software portfolio as cloud services; 2) it has found a way to bridge the gap between VMware virtualized environments and OpenStack clouds; and, 3) it has partnered even more closely with mobile development community, strengthening its server-to-mobile infrastructure with support for SWIFT development and forming a strategic relationship with GITHUB for better hybrid cloud support.
To me, cognitively-enabling the IBM software portfolio is the most momentous of all the changes taking place at IBM. What I found when touring the EXPO floor was several IBM software organizations that were actively blending their traditional software offerings with IBM Watson and analytics products. My analysis: the result of this blending has the potential to be a true game-changer for IBM in certain disciplines such as systems, storage, network, application, database and operations management.
Allow me to explain:
- For decades, vendors have built management products that capture vast amounts of data, placing that data in log files and in databases where humans can examine that data looking for problems or ways to better configure systems to improve performance.
- The problems with these management products are that they require advanced skills and they are labor-intensive. For instance, it could take a team of people from systems, storage, application and database disciplines to figure out the root cause of a performance or failure problem. These people need to comb through mountains of data trying to isolate problems (and sometimes finger-pointing occurs).
- Now consider using system intelligence and analytics to tackle this problem. Imagine using IBM Watson analytics algorithms to sift through mountains of Big Data to look to find the root cause of a problem. Using analytics, the time it takes to identify a problem can be greatly shortened, leading to greatly-reduced human labor costs as well as improved Quality-of-Service (QoS).
- In addition to using analytics for root cause analysis, it can also be used to examine application/database behavior, helping IT managers better plan their resource usage. Workloads could be better balanced and executed more efficiently – again, improving QoS.
- Further, the sequences used to solve problems can be recorded in a knowledge base, helping successive generations of IT managers learn how certain issues were resolved from a historical basis.
- Finally, consider the concept of machine learning. A system could run through the sequence of identifying a problem – and point an operator toward a solution. That operator could then decide to automate that solution, so – should the problem reoccur – the machine could be given permission to automatically take corrective action.
In short, by “cognitively-enabling” applications, machines will be playing a greater role in managing themselves.
Clabby Analytics has discussed cognitively-enabled/analytics-driven solutions in greater depth in this report on log file analysis; and in this report on the future of distributed systems and mainframe management.
But what changed at InterConnect was that there is now a general awareness across IBM about the importance of blending traditional solutions with cognitive systems.
More and more IBM disciplines have come to recognize that analytics and Watson will change the way their products will behave in the future, introducing new insights, streamlining performance, simplifying operator interaction with underlying systems and subsystem while improving efficiency and QoS. Cognitive computing is starting to take a huge step forward into the application and management mainstream.
The above example described how the traditional IT Operations Management (ITOP) market is changing, moving to a new analytics-enabled level known as IT Operations Analytics (ITOA). But I found several other examples of cognitive application development taking place at InterConnect. For instance, the Tivoli group would like to replace certain dashboard driven management packages with a cognitively-enabled, voice driven, Watson-like interface that business executives could use to better monitor the state of their systems and business operations. Plus, the head of IBM’s collaborative software organization told me that cognitively-enabling her software portfolio was a high priority.
One final point on cognitively-enabled applications. IBM is in a unique position in the market because no other vendor offers a technological-equivalent to the company’s cognitive, natural-language-driven Watson computing environment. Although IBM software organizations are starting to understand the importance of cognitively-enabling their applications, I could find no evidence that IBM’s Watson organization is reaching out to its partner ecosystem encouraging partners to mesh their management products with Watson algorithms.
I found few Watson resources at InterConnect – but I know I’ll find an entire cadre of Watson people at the next big event that I attend – IBM Insight in October. My key point here is that cognitively enabling applications is going to have a major impact on a lot of software products – and IBM channel partners and certain independent software vendors (ISVs) are going to have to go through IBM to fully achieve cognitive enablement of their own solutions. I see IBM starting to cognitively-enable its own portfolio – but I’m not yet seeing IBM reach out to assist its partners in the same process.
The Cloud Announcements
The big cloud announcements at InterConnect included IBM’s statement that it had cloud-enabled all of its leading applications; that VMware and OpenStack are working together; and, that IBM has tightened smoothed the development of mobile/hybrid cloud applications with better support for Apple’s SWIFT language and GITHUB’s enterprise services.
The IBM statement that it had cloud-enabled all of its major software packages came as no surprise – the company has been working for years to offer its major programs as cloud services. What is important about this announcement is the sheer volume of customers that can now get access to cloud enabled IBM applications. For instance, IBM’s WebSphere alone has 100,000 customers and runs over 200,000 instances. Cloud-enabled applications should give IBM customers new deployment options – but it should also open new opportunities with new customers for IBM’s existing portfolio.
On IBM’s announcement that it had formed a strategic relationship with VMware to unite virtual machines based on its solutions with the IBM cloud (an OpenStack cloud environment), I see this as great news for the existing VMware base. Why? Because it means that VMware users will not have to abandon their current, preferred solutions to become part of the industry’s budding open source cloud environment. My read on this is that VMware has recognized that basic virtualization is becoming standardized and hence is willing to become part of a open source virtualization layer – but the company will continue to add compelling infrastructure and management value.
IBM’s increased support for GitHub and SWIFT shows the company paving the way for better hybrid cloud integration, as well as improved access by mobile devices to hybrid cloud services. The benefits of IBM’s strategic relationship with GitHub is that the GitHub community (12 million developers) will be able to link its applications with IBM’s cloud environment without having to deal with a lot of infrastructure complexity (GitHub will interface with the IBM cloud through IBM’s Bluemix – a set of API solutions that simplify application deployment). When linked with the IBM cloud, GitHub developers will be able to take advantage of code in IBM’s portfolio to build new, hybrid cloud applications (including applications that take advantage of IBM’s cognitive capabilities as described earlier in this report).
Finally, SWIFT is a development language supported by Apple that has the support of over 11 million developers. Increased SWIFT support enables IBM to open the door for more mobile-to-backend applications – further expanding the number of applications that IBM services can support.
IBM runs four major customer events each year in the United States: PartnerWorld, Edge, Insight and InterConnect. PartnerWorld serves IBM business partners; Edge is primarily for technologists; Insight focuses on analytics and cognitive computing; and IBM InterConnect focuses on mobile, cloud, application development and security. Starting last year, IBM opted to combine three previously separate conferences (the IBM Impact, Innovate and Pulse conferences) into Interconnect. By combining these events, InterConnect has now become a massive gathering that includes over 22,000 customers, business partners, consultants, press and analysts, and IBMers. The conference features hundreds of customer led case studies; specialized training, strategy and product sessions; and a large EXPO center where vendors and business partners display their products and describe their services.
As I observed last year, although the conference is nominally focused on mobile and cloud, attendees can easily find access to a wealth of non-cloud, non-mobile technologies including access to sessions and demonstrations on hardware, analytics, DevOps, the Internet-of-Things, security and so on. With this broad selection of technology sessions and demos, attendees can pretty much tailor the conference to their specific needs.
On the EXPO floor, I found that the cloud organization claimed cognition with analytics cloud offerings and a new Cloudant-based data service, with analytics for Apache Spark, with dashDB, with BigInsights on Cloud, and so on. The mobile organization demonstrated how cognitive computing services can be delivered to a wide variety of mobile devices. The Internet-of-Things organization showed how analytics is used to analyze the activities of instruments. No matter which booth an attendee visited, or which session an attendee attended, cognitive themes permeated the event.
My one big concern about IBM’s cognitive computing message is this: I saw lots of activity taking place within IBM’s software organization to find ways to cognitively-enable traditional workloads. But it is not clear to me that IBM is paving the way for its own partner ecosystem to cognitively-enable their respective applications. I’d like to see more progress on this front – and will be looking for clearer cognitive-enablement direction for 3rd parties at the next big IBM event at Insight in October.