Most Read This Week
Cloud Computing Viewpoint
The Future of Cloud Computing: Part 1 - VDI is DOA
The answer is not SaaS, nor VDI, nor Cloud; rather, an evolutionary compilation of all these technologies
By: Ryan Hughes
Aug. 20, 2012 06:00 AM
The impact that Cloud Computing has brought to the IT industry to date has been primarily beneficial to application developers, system admins, and network architects, and not directly to end-users of technology.
Yes, IT developers and architects leverage cloud computing’s flexible and virtualized compute, storage, and network infrastructure to build resilient applications that eventually benefit end users due to improvements in speed-to-market and improved up-time statistics, but the direct benefits to the tech-needy end user are still rarely recognized.
Most daily users of personal and business class applications don’t have the turnkey, on-demand access to the applications they need. At work, their IT departments at work are too slow in delivering the apps they need or refuse to provide them due to cost, limited resources, or lack of recognized need. At home, users struggle to deploy software themselves due to complexity, time involved, or again, cost.
However, advances in cloud-powered software and service delivery have started to revolutionize the way that end-users (both business and consumer-level) think about acquiring the tools they need to succeed. These innovations will finally give end-users with their piece of cloud computing value and change the way software is delivered, licensed, and used both on-line and off-line. Over the next several weeks, I will be releasing several blog posts on the topic of the "Future of Cloud Computing". Below is Part 1, which describes the unrealized promise and eventual demise of virtual desktops.
Innovations in streaming application code… rather than streaming pixels… will kill VDI before it even fully arrives.
Do users really like or want Virtual Desktops? From the start, the concept of virtual desktop infrastructure (VDI) is flawed for most real-world applications and use-cases. No matter how optimized VDI compression companies claim their proprietary algorithms might be, they are still trying to push a proverbial “watermelon of pixels” though a relatively pinhole-size network to get what you need to your device. It almost seems like all the stars have to align before VDI actually works for the every day, multi-location worker.
VDI technology refresher
Virtual Desktop Infrastructure (VDI) is a method of enabling end-users with a client device (PC, laptop, tablet, etc.) to access, log-in, and utilize a remotely hosted desktop environment. In order for you to get access to and interact with the remote environment, compressed screen shots of the display (what you would see if you were standing in front of a monitor) of the VDI instance are streamed continuously via network connection to your client device’s display screen. Meaning, that a user could have access to a completely different environment including OS, applications, and network without actually having that environment installed on their physical client device.
What’s so wrong with VDI now?
For the typical everyday business user, who works from a combination of office, home, client-site and car, using a virtual desktop sounds perfect, but in actual practice, it’s a real productivity killer due to several key flaws.
How were we convinced streaming screenshots was "the right way" anyhow?
Undoubtedly there are benefits of VDI, but most of the benefits are to the IT staff, not end-users. Most of these benefits to the IT staff surround topics of license management, patching, and security. Although I understand these benefits, I don't know how IT shops got on the path of streaming pixels with VDI rather serving the code instead which would allow them to better optimize and control application delivery and licensing than what streaming screenshots could.
Using the server-side to deliver application functionality, data, and licensing on-demand to devices directly
Sending pieces of the code to your device, using your local device’s compute processing to run it, and then getting updates pushed from the mothership server whenever you connect or security requires it seems like a much more streamlined approach to a VDI-like environment than relying on a high-speed connection to stream pictures of screenshots from a remote data center slice of a server. In this scenario, IT admins still get all the manageability benefits and licensing controls for deploying applications on-demand that they get from VDI… all without spinning up an entire cloud infrastructure to host a VDI backend and without wasting perfectly good client-side resources.
How to replace VDI... Streaming application code, not pixels
The next step - making "Cloud-bursting" workloads a reality
More Advantages of streaming application code rather than pixels
Although the technology to pull off this type of code-streaming environment might not be full baked yet, the groundwork for replacing pixel-streaming VDI has already been laid. As the cost drops for cutting edge client devices and their amazing processing and graphics capabilities continue to wow customers and set expectations on user experience, VDI implementations will continue fail at achieving their once great promise to stream any application to every user via only a web connection. It seems that VDI is perhaps only a patch-over solution while we wait on something better to come about. Code streaming to client devices may be that answer.Watch for my upcoming post: The Future of Cloud Computing - Part 2: Why PaaS will fail and how Software-Stacks-as-a-Service (SSaaS) will replace it.
Reader Feedback: Page 1 of 1
Subscribe to the World's Most Powerful Newsletters
Today's Top Reads