For as long as I can remember, large organisations have been trying to reduce the cost of supporting desktop computers. It was cheaper and easier ‘back in the day’ when the mainframe computer supported a bunch of dumb terminals which had no local processing power, a text-only green screen and every processing request was sent up and down the network for the central processor to time-slice the work, running all the applications for all the users.
We’re not going back to that. Or are we? ‘Desktop virtualization’ is becoming more popular. Why support a bunch of individually installed PC’s operated meddlesome, technically insecure users when there’s the option of hosting all those desktops on a server in the data-center, to centrally secure and manage your data under centrally registered and deployed applications…?
The desktop machine in this scenario needs be just a low-spec device capable of redrawing the screen whilst all the real work happens on the server. The user’s desktop becomes a ring-fenced, tamper-proof and transient asset. At a stroke you minimize the footprint and management costs of the device.
This is the intended use of Windows Thin PC (WinTPC) product, which is a locked down version of Windows 7, designed for customers to re-purpose existing PCs as thin clients using VDI – Virtual Desktop Infrastructure.
There’s a lot of new and old terminology washing around here. Thick client, thin client, zero client; there are some key differences:
- Think of the PC as a Thick Client (the PC, not the user – but then again…); yes it uses network resources such as directories, mail servers, storage and printers; but it runs it’s standard office productivity applications locally from it’s own hard disk, Windows, registry and all. This may be the industry-standard way of doing things but it is also the most expensive.
- Zero Clients: A new form factor, the Zero Client is a thin client terminal with no local operating system or storage, and depends on the server for all of its computing and transaction capabilities.
- Hardware Configuration: Traditional thin clients differ from a PC in that they have a lower CPU and less RAM than a typical PC, need a smaller power supply so typically use less energy and have little (if any) local storage for data and applications.
- Functionality: Traditional and Zero client devices must always be network connected since they cannot run local applications.
- Management: thin clients may require less management than a PC but they are not “zero management” since they still need firmware and security updates.
- Cost: The thin client may be cheaper but it’s not free. There’s a purchase cost similar to a low-spec PC. In these financially straightened times, a low capital spend may appeal to the Finance department. For thin clients and zero clients, you will need software licenses for VDI to allow multiple users to access applications on the server.
- Multimedia capabilities: You may or may not need sound and video or advanced graphics on your remote desktops, which is just as well, as not all thin clients or zero clients have built-in multimedia capability. While this keeps the cost down, this limits the use of the device.
Some organisations I have worked in have gone for specialist devices such as HP and Wyse ‘terminals.’ These were originally up-rated mainframe terminals, then moved onto running slimmed down versions of Windows XP, but now use Windows Embedded to create low footprint thin clients compatible with Microsoft applications.
With Microsoft pushing Windows Embedded and Thin PC in conjunction with their software management tools and Software Assurance (a glorified licensing register and administration service) it looks like a lot of older PC’s are about to get an extended life as remote desktop clients. AJS