IBM is the original, and in some areas still, pre-eminent enterprise IT company, but since squandering its lead after launching the PC era, it seems to continually be playing catch-up to nascent IT trends and markets.
Although IBM pioneered the rental of computing resources in the 1960s and was early to offer modern cloud services, it remains a second-tier player behind the big-two, AWS and Azure.
Likely a combination of its size, history, internal politics and corporate inertia, whatever the reason, IBM didn't have that 'betting-the-company' commitment to cloud computing despite quarterly management bromides and several significant acquisitions.
After a decade of almost continually declining revenue that has shaved more than 30% off of the company's top line, IBM appears ready to focus on a hybrid cloud future.
There are three catalysts for IBM's rejuvenation:
- The 2019 acquisition of Red Hat, which brought cloud software, market and developer credibility and a fresh, aggressive attitude to IBM's cloud efforts.
- Last year's retirement of longtime CEO Ginni Rometty, who mouthed the right words about cloud and AI, but never galvanized the company into streamlining processes and delivering the necessary products.
- The impending spin out of its enormous, but legacy IT services business, aka NewCo. In announcing the spinoff last fall, new IBM CEO Arvind Krisha spelled out the hybrid cloud imperative:
IBM is laser-focused on the $1 trillion hybrid cloud opportunity. Client buying needs for application and infrastructure services are diverging, while adoption of our hybrid cloud platform is accelerating. Now is the right time to create two market-leading companies focused on what they do best. IBM will focus on its open hybrid cloud platform and AI capabilities.
IBM contends that the industry remains in the early stages of cloud evolution with 80% of its projected trillion-dollar TAM yet to be unlocked, first by re-factoring and containerizing enterprise applications and then fully redesigning them for multi/hybrid cloud environments.
IBM multi-cloud built on Red Hat and OpenShift
Application containers are central to IBM's cloud strategy, however, in contrast to Google, whose enterprise playbook is built on the same technology, IBM doesn't lead with public cloud services, but rather on-premises infrastructure. Furthermore — and here we see Red Hat's influence — IBM contends that building hybrid clouds with eminently portable open source software provides the most flexibility, cost-efficiency and security for enterprises looking to balance the convenience, scalability and technological innovation of public cloud infrastructure with the performance, regulatory control and cost efficiency of on-premises infrastructure.
To support its open-source-based, on-premises-first cloud strategy, IBM cites results from a survey it sponsored of IT "decision-makers" around the world in which:
- 89% agreed with the statement that "open source allows for a more open and flexible hybrid cloud strategy."
- 83% agreed that "A hybrid cloud IT infrastructure leverages open source for greater efficiency and scalability in the future."
After the events of last year, these same survey respondents showed a marked increase in the share citing data residency, regulatory and security requirements as reasons for maintaining infrastructure outside of public clouds.
Red Hat and OpenShift undergird several elements of IBM's hybrid cloud strategy, notably by using Linux across the infrastructure layer, OpenShift container environments at the application level and Cloud Pak containerized AI applications and the Red Hat marketplace for cloud software and development platforms.
Linux as the glue unifying IBM's disparate hardware platforms
Perhaps IBM's most significant competitive advantage is the massive installed base of IBM hardware, particularly those ostensibly obsolete dinosaurs of the cloud era, mainframes. Indeed, more than 70% of large corporations and virtually all multinational financial institutions and large retailers use IBM z System mainframes. According to Tom Rosamilia, SVP of IBM's Systems organization, mainframes handle two-thirds of enterprise mission-critical workloads, but account for only six percent of total IT spending. Even assuming that IBM is sandbagging the numbers, the fact remains that mainframes are efficient, mature, reliable and well-understood multi-tenant systems, making them ideal private cloud platforms.
IBM has positioned both its z and Power (RISC processor) Systems for the cloud by releasing Linux and OpenShift versions for each and allowing x86 workloads to
run on z System logical partitions (LPARs) as a guest within a z/VM or KVM hypervisor. Its LinuxONE III product tightens the integration via microcode added to the standard z System hardware to allow native communication between a Linux kernel that the mainframe's underlying cores, memory and I/O devices. In describing how to install OpenShift on z-series hardware, Red Hat makes a compelling case for using mainframes as an enterprise container platform, citing:
scalability that can grow to thousands of Linux guests and millions of containers, and vertical scalability that can non-disruptively grow databases without requiring any partitioning or sharding while maintaining response times and throughput.
Rosamilia touted the IBM Z hardware and its GBS hybrid cloud services in a webcast last week claiming that the combination helped organizations significantly cut the time-to-market for new applications, reduced risk by using mature hardware assets and proven methodologies and cut the TCO of app modernization efforts by up to 30%.
IBM sees AI and data analysis applications as an ideal target for hybrid cloud environments and has made several recent moves to facilitate application development and modernization, including:
- A partnership with Palantir to develop a new application, IBM Cloud Pak for Data, that combines IBM's Watson AI technology with Palantir's Foundry data analysis product. As part of the partnership, Palantir has adopted OpenShift for hybrid cloud applications.
- Acquired Instana, an application performance monitoring and observability company to improve IBM's cloud management and automation capabilities by feeding Instana-derived data to Watson AIOPs.
- Introduced new IBM Wazi Analyze features to integrate z Systems into DevOps automation pipelines and ease the transition to mainframe systems. The additions speed development by providing granular code analysis and dependency mapping between different components.
As the historical chart of revenue illustrates, IBM remains a company in decline and has turned more corners than men in an Escher painting, however its Q4 2020 earnings showed signs of progress, namely:
- 20% annual growth in cloud revenue
- 17% growth at Red Hat with more than 2,800 customers for its hybrid cloud platform
- A strong mainframe refresh cycle following the 2019 introduction of the z15 leading to significant margin expansion in its systems unit and installed IBM Z compute capacity "3.5 times the level of a decade ago, with 60% of our installed base now in new workload areas like Linux," according to IBM's CFO on the Q4 earnings call.
IBM's coherent software strategy that spans mainframes, RISC (Power) servers and third-party x86 boxes provides organizations with an unrivalled range of capacity, reliability and cost options.
With Red Hat and its cloud services, IBM has a compelling software and services substrate for hybrid cloud environments and one that rivals VMware (Cloud Foundation), Microsoft (Azure, Azure Stack), AWS (with Outposts) and Google Cloud (Anthos). The cherry on top is the availability of Linux and OpenShift on IBM Z and Power Systems, which could turn its legacy systems business into a solid foundation for on-premises hybrid environments. It will be interesting to see how IBM navigates the NewCo divestiture and whether customers treat a re-focused Big Blue as a strategic, first-tier cloud vendor.