Summary Bullets:

• The nature of the problems changed from local to network, requiring expensive network support rather than using PC technicians.

• It’s important to remember that many concepts in modern enterprise IT come up again and again cyclically, only to be characterized as ‘paradigm-changing’ by enthusiastic marketing teams.

The idea of remote desktops or streaming an employee’s Microsoft Windows desktop to a more affordable device or even in a window on a remote employee’s self-owned hardware has been around for a long time. The reasons were simple – less spending on desktop computers, easier support both on-site and particularly for remote, better security, and remote desktops worked particularly well for temporary or project-based access for outsiders. Citrix and VMware both made hay with remote desktops, especially in the era where PC hardware and Microsoft Windows itself were considerably less reliable than today.

But there has always been a problem – the experience has always sucked for the end user. PC prices for standard corporate desktops have declined over the years, making the savings that those thin client hardware devices were supposed to bring vs. PCs to be minimal. Thin clients also didn’t help with other end-user hardware issues – like monitors, keyboards, or mice. Worse, it turns out to be way more expensive to buy server hardware, support and maintenance to host virtual PCs than it is to just buy PCs for the workforce. Lastly, software incompatibility, network lag, and the disenfranchisement of the end user all meant that calls to the help desk either did not decrease or increased. At best, the nature of the problems changed from local to network, requiring expensive network support rather than using PC technicians. Plus, the company had to pay for the remote PC software and likely had to pay for more bandwidth to support it.

Yet every few years, the idea of essentially returning end users to a mainframe-style environment with what amounts to “dumb” terminals reappears. To be clear, there are good use cases for remote desktops. Remote environments, and jobs where third parties need access to the corporate desktop for the duration of a project, is in particular a great use case. Harsh environments, or environments where no corporate data can remain local, such as home-based call centers, are also a great use case. But it never came close to being the default way of accessing the desktop in the enterprise, despite what followers of remote desktop technology claimed. It was always said that the fastest way to kill a remote desktop environment was to force VPs or C-suite occupants onto remote desktops. The inconsistent end user experience and limitations imposed by the remote PC environment always led to execs demanding their laptops back.

This little trip down memory lane isn’t just random. HP, (not HPE) recently announced it is retiring its HP Anywhere remote desktop solution in May 2026, with support for customers running until the last day of October 2029. HP’s solution was from its Teradici acquisition in 2021 and ran on the once-promising PCoIP (PC over IP) protocol. Citrix is a shell of its former self and VMware sold the entire Horizon and Workspace ONE business to Omnissa, a PE backed company, in 2024. Microsoft with its Microsoft Azure Virtual Desktops, and Amazon WorkSpaces are the latest companies to revive this model, but this time it’s run from the cloud. These latest iterations are, like past remote desktop/VDI solutions, still only really useful in niche use cases, despite enthusiasm from Amazon Web Services (AWS) and Microsoft.

It’s important to remember that many concepts in modern enterprise IT come up again and again cyclically, only to be characterized as ‘paradigm-changing’ by enthusiastic marketing teams. That doesn’t mean old concepts in a new suit won’t work – but it does mean IT professionals who pay attention can avoid the easy mistakes from the last cycle.