Liquidware Bets On Cloud-Based Virtual Desktops
Liquidware is betting that cloud-based virtual desktops will sprawl out in a similar manner to virtual servers, and is partnering up with the major cloud vendors in anticipation.
The early promise of VDI hasn’t really played out in practice, and pretty much all of the analysts ended up revising their now obviously optimistic predictions of 40% or more of desktops becoming virtual to more modest estimates of 10-to-20%.
“A lot of early adopters ran with scissors and hurt themselves,” says Liquidware co-founder Tyler Rohrer. They discovered that the server, storage, and network demands of virtual desktops were substantial, completely eclipsing any imagined savings of a reduction in the number of desktop PCs required. Thin clients never really took off, and BYOD tablet devices could only run native apps or browser-based software.
However, technology continued to improve. The capabilities of modern flash-based storage, multi-gigabit, low-latency networks, and the impressive compute density of today’s CPUs makes the prospect of running simple virtual desktop environments seem fairly trivial by comparison.
The hype has also fallen away. It seems that VDI has now passed through the trough of disillusionment and people are getting on with using it productively. At least to achieve relatively simple tasks. “With VDI, the easy stuff has already been done,” says Rohrer.
“You know our history in VDI,” said Rohrer, “and now the same thing is playing itself out in the cloud market.” Liquidware announced support for AWS Workspaces at AWS re:Invent in November 2017. “While I can’t say too much, it’s safe to say that we’re also beginning our collaboration with Google and Microsoft, which is really exciting.” Expect to see Liquidware support for Azure and Google Cloud Platform in the fairly near future.
Personally, I’ve always been a little cool on the idea of virtual desktops, cloud based or not. I admit I had a brief love affair with some Sun Ray 1 thin client systems that would resume my session as soon as I inserted my smartcard, but alas the early promise never translated into a solution that was more broadly useful.
I challenged Rohrer that virtual desktop seemed to remain a somewhat niche proposition.
“It was both a blessing and a curse that happened about 30 years ago which was this little thing called Windows,” said Rohrer. “It was a blessing because back then we had a standard, and everyone could develop software to that standard.
“But because of that, we ended up getting a billion seats of Windows out there, and six million software titles,” he said. “Without those six million software titles, the migration to cloud would happen a lot, lot quicker.”
Rohrer makes a good point. Writing software is about 80% putting bugs in, and then another 80% taking them out again. Rewriting all that software isn’t going to happen overnight, and we saw the same issue play out when VMware first became really popular. Physical-to-Virtual lift-and-shift moves were *extremely* common, and I personally know of more than a few aging NT systems that were given a new lease on life by being taken off aging hardware and run as a VM.
The same is undoubtedly true of desktop based software titles, and the vital business functions they support. The software needs to run *somewhere*.
“It’s something we need to contend with,” says Rohrer. “I agree, on-premises VDI is a very limited niche case, maybe 12-15% of your total PC population.”
“What does go away with cloud-hosted desktops is hardware complexity,” says Rohrer. “The challenge of managing Windows at scale doesn’t go away.” And that’s where Rohrer sees Liquidware shining. We have seen a tremendous example of Jevon’s Paradox where the ease with which new VMs can be created has encouraged an explosion of VMs.
So it is with containers, and so it will be with desktops when it is simple and easy to simply spin up a new one. Managing this sprawl then becomes a higher order problem that organizations will need to solve, and just as we are seeing services created to manage these sprawl issues for servers, Liquidware is betting that we will see the same sprawl management challenges with desktops.
For some people—such as the poor souls trapped in Australia without access to high-quality broadband connections to the cloud, like myself—virtual desktops are likely to remain promising but impractical. But for others, and their number increases daily, the base resources required are easily accessible, and for the virtual desktops may make sense for a lot of situations.
We essentially have two competing rates of change. On the one hand, we have the rate of applications being rewritten for cloud, or simply abandoned as the work is moved to entirely new ways to achieve the same result. On the other, we have the increasing ease with which these applications can be moved to virtual environments where they continue to be useful.
If the first change happens quicker than the second, the need for virtual desktops will go away faster than Liquidware expects. But if change happens more slowly than the techno-utopians would have us believe, there could well be plenty of money to be made while Liquidware builds new products to manage the new problems of working with desktop style apps in the future.
Grab your popcorn.
This article first appeared in Forbes.com here.