Commentary: There’s no Room for Sentimentality in Enterprise IT

As Broadcom takes a knife to VMware, cutting services, support and staff, the IT community has expressed dismay and sadness for technology that, for many, represents a significant part of their career. However, history is littered with sentimentality for technology that serves a single purpose – delivering business advantage. It’s time to move on.

Background

Anyone with a reasonably long history in the IT world will have experienced many transitions over their career. Typically, these sea changes occur perhaps once in a decade. In the 1970s and 1980s, the world revolved around the IBM mainframe, which was the gold standard for enterprise computing. A Stanford Computer Science class claims IBM owned perhaps 70% of the computer market, while I’ve seen suggestions that in the late 1970s, this figure was as high as 95% of the global market TAM.

In the mid-1980s, the personal computer started to gain a foothold (again, driven by IBM) and by the 1990s, the minicomputer, powered by UNIX, saw a major revival with solutions from Sun Microsystems, Sequent, HP and IBM. The 1980s and 1990s saw diversification into many computing models, including departmental and desktop systems. Partly, this change was driven by IBM not releasing a new mainframe platform between the System/370 in 1970 and the ESA/370 and 390 architectures in 1988. (Side Note: the delay in releasing a new mainframe architecture is blamed partly on the development of an internal RISC solution and also on IBM anti-trust litigation).

As Windows developed and evolved into a server platform, the x86 architecture started a period of dominance (driven by an aggressive Microsoft), which entered a symbiotic relationship with VMware. Competing x86 ISAs (such as Itanium) fell away (as has SPARC and PA-RISC), leaving Intel with free-reign and, of course, for x86 UNIX (otherwise known as Linux) to develop and become the dominant non-Windows architecture in the data centre.

VMware dominated on-premises data centres during the 2000s and 2010s, while that plucky upstart, Cloud Computing, started to get a foothold and transform itself into the default deployment model for the majority of modern enterprises. Nearly 20 years on, the public cloud is a multi-billion-dollar business showing few signs of slowing down.

Yet even as we hail utility computing as the model of the future, the technology world is changing again, with AI the current shiny new toy that everyone wants to be using. There’s no doubt that Shiny Object Syndrome is alive and well in the technology world.

Evolution

Few industries in the business world evolve at the same pace as technology. If we examine Finance, Healthcare, Fashion, Manufacturing, Logistics, Aerospace and more, we see mostly evolution, but rarely do we experience massive shifts in practices or capabilities. Generally, most industries change significantly over the course of decades, whereas the technology world can change in a matter of weeks.

Using the emergence of Generative AI as an example, within a year of the launch of ChatGPT, we saw the release of new large-language models (LLMs) from all the hyper-scale platforms. The industry is already talking about artificial general intelligence as a near-term possibility, while OpenAI has recently released a text-to-video LLM that creates photorealistic video clips. AI is at the forefront of driving the next wave of technology and is currently where the big money is headed.

Forwards not Backwards

Unlike some industries, like architecture or fashion, technology rarely goes retro. We will never return to the mainframe era of the 1970s because modern hardware is orders of magnitude faster and cheaper. However, we can learn from some mainframe principles and, of course, operational processes, which by their nature had to be more efficient in the 1970s and 1980s due to the relative cost of computing.

Like it or not, the technology industry moves forward at a rapid pace. While little may have changed in the space of 12 months, look back five or ten years, and the advances are enormous. There is no room for sentimentality in IT, where one technology architecture can be rapidly subsumed by another in a few short years.

Server virtualisation fits this pattern and has rapidly become part of the computing ecosystem. While VMware has been the most successful on-premises example of this technology, the public cloud doesn’t run on VMware ESXi because the cost and operational model would be too onerous. (VMware vSphere is supported by cloud platforms but is probably statistically insignificant in revenue terms.)

Rumours of my Death

As Mark Twain is frequently quoted as saying in 1897, “The reports of my death are greatly exaggerated”. The same can be said for the mainframe, tape, hard disk drives, server virtualisation and many other enterprise technologies. Most technology doesn’t disappear overnight, although that may be the assumption when legacy technologies are no longer the “tech du jour”. IBM’s mainframe (zSystems) is still a multi-billion-dollar business. Tape is seeing a resurgence, while Japan has only recently relaxed the rules on submitting government submissions on floppy disks. Most widely adopted technology lasts for decades, with a long tail of replacement or refresh. It’s clear that VMware will not disappear overnight but will not exist with the ascendent success it had in the 2010s.

Sentimentality

With so many big bets made on technology, including rapid adoption, it’s easy to see how technology professionals become so personally engaged with new trends, platforms, and solutions. At the outset of a new technology wave, few people understand the requirements, and there are career-defining moments to be made (as well as a lot of money). When the gravy train stops rolling, people are understandably upset. However, it’s not just a financial story. Long-lived careers with a single technology or vendor also create personal memories and experiences that stay with us for many years.

The Architect’s View®

From a business perspective, technology is a tool. All technology forms one of two purposes – to either make a business more efficient or to create business advantage. AI is certainly in the latter category, while VMware started as an efficiency play and rapidly developed into a business agility solution.

In IT, being able to work on a single solution for one or two decades is a luxury that doesn’t come around often. I’ve worked on mainframes, open systems platforms and shared (SAN) storage during my career. Each has waxed and waned over time, requiring a reskilling phase as the next wave of technology emerges.

Technology professionals and businesses need to be less sentimental about technology platforms and more hard-headed on what delivers the best business value. VMware is probably in the decline and maintenance phase of technology that we saw from the mainframe in the 1990s. But in reality, this process is perhaps 5-10 years in, firstly from the adoption of the public cloud and second from the transformation of applications towards containerisation.

VMware was a great company, but every dog has its day. It’s time to move on. Let it go and transform your business to the next technology that delivers even more business advantages.