Thin Clients Are Back--It's About Time

Initially, the thin client effort was created to take the PC market away from the existing companies that dominated it. As a result, neither Intel nor Microsoft backed it.

But over the last decade, things have changed. Both Sun's and Oracle's thin client efforts failed, and Dell and HP took the market over quietly through acquisition. Recently, two Intel competitors jumped in this market with both feet, and Microsoft's biggest competitor, Google, developed a platform for it. Even Microsoft itself--through offerings like Office 365--appears to be embracing a cloud-delivery model consistent with thin clients.

Suddenly, we actually are ramping up to a thin client model.

Interesting thing though, it isn't IT leading the way--it is computer gaming and tablets, which may be the perfect thin client hardware for the near term.

While a number of cloud gaming companies have come to market after OnLive's launch, OnLive was the first to demonstrate that you could get the necessary performance to play a first-person shooter that was hosted remotely. They were also the first to offer the option of running other applications like Microsoft Office as a service.

Since then, companies like Mainframe2 have positioned themselves to provide workstation-class applications (Adobe to CAD) from a hosted environment. This has eliminated much of the cost of deploying a thin client solution, which used to require a massive investment in networking hardware prior to beginning an even more expensive installation of client and server hardware.

Because gaming solutions have to run on existing telephone and cable networks and use either low-cost (under $100) or existing set top boxes or tablets. The result is that the upfront cost has largely been eliminated.

At back-to-back events, both AMD and NVIDIA announced hardware to host high-performance desktop computing from the cloud. Currently NVIDIA's VCA Appliance stands out as uniquely well designed for this new use.

Initially, most companies focused on supporting the client, but still used traditional servers or PC blades to provide the services. The first simply didn't have the I/O needed to provide a cost effective hosted desktop solution. And the second didn't scale particularly well, largely due to a combination of low volume and a lack of R&D needed to create the specialized hardware which could dynamically shift loads based on use.

In short, the last missing link was to have the at least one of the PC processor vendors being designing a solution to meet this need. Fortunately, both NVIDIA and AMD stepped up.

The first thin client efforts typically required client hardware that was as big as, and often more expensive than, the PCs they were designed to replace. The argument back then was that the savings in support would justify the massive additional cost. IT didn't buy that argument for a ton of reasons.

Then, even though the hardware got far cheaper, so did PCs. And employees became mobile, making the old solutions, which generally required wired connections, even less attractive.

But these latest efforts are designed to work with tablets, which increasingly have USB and HDMI ports for better control and larger screens. Suddenly there is a client that can scale from the mobile world to the desktop, putting in place the final ingredients to our recipe for next generation computing.