Thursday, 29 April 2010

Desktop Virtualisation - Fully baked or a damp squib?

Silicon.com had a piece recently about desktop virtualisation, apparently, it's all going to kick off this year. Is it? Is it really?!

In my view, it's a long way off and I'll explain why:

We have off-shore facilities that need access to our main CRM platform and historically another part of the group has provided this functionality using Virtual Desktops. We also have off-shore developer capability that also use Virtual Desktops.

I'm currently looking to take over this functionality, and having looked at the pain that it produces, the last thing I want to do is replicate it. It seems to be continuously unreliable, it is slow, fails at inopportune moments and usually does a great job of over-promising and under-delivering and is staggeringly expensive to expand or upgrade.

When I tell people the price for additional machines, they look at me as though I'm mad and I don't blame them: £600 for a basic virtual machine and £1,200 for a developer standard machine. The usual response I get is that they could go down to the local PC store and buy a machine for a lot less than that. And I wouldn't blame them!

How can virtual desktops get lift-off when the price point still isn't right. Surely a virtual desktop needs to cost less than £50 for a good spec virtual PC before we can really say Virtual Desktops have arrived. Anything else will just not make financial sense.

Also, in general so many things can get delivered by browser nowadays, the need for virtual machines is diminishing anyway. A basic PC with Windows and a browser costs so little nowadays, why would I want a virtual machine and all the hassle associated with them.

As for my remote developers, the best thing I can do for them is use a VPN or MPLS connection out to their office and create a LAN extension giving them direct access to our development boxes. Job done, lots of money saved.

So, why are Virtual Desktops supposed to go big this year. I've no idea, I've really no idea.

No comments: