Unlike earlier efforts that provided either low performance or a dedicated one-to-one, server-to-desktop system, the key to hardware built on the Kepler processor is that it is designed to auto-scale to the required performance level. Moving dynamically from a shared resource to dedicated resources based on processing loads that range from light email to full-on CAD or media editing-and displayed on everything from traditional desktop monitors to iPads and smart TVs-could be the ideal solution for IT managers struggling to meet the conflicting needs of a BYOD world while still providing a consistent level of breadth and security.
Consider the alternative. Amazon Web Services are covertly making their way into enterprises, as employees are increasingly willing to pay for them in order to obtain quick solutions to common storage, database and networking problems, to name three. At recent EMC and CIO events I've attended, this use of credit cards to bypass IT was listed as a trend that borders on an epidemic. If IT organizations don't act fast, though, employees will continue to invest in third-party services in the absence of official IT policy.
This new class of Kepler-based servers will be coming into enterprises and technology-intensive small businesses. The aforementioned services from Amazon and other vendors will likely enter these companies where these servers don't exist, but the need for BYOD coverage does. Right now, that's majority of businesses.
The not-so-subtle message, then, is that IT needs to get its arms around both the opportunity and the risk of technologies such as Kepler and provide alternatives to anticipated Amazon and Amazon-like services. Otherwise, IT will find itself out of the loop for yet another layer of end-user technology-and, since budgets typically follow usage, this means IT budgets will suffer as a result.
We are clearly moving to PCs-as-a-service. Like any major change, this one promises less-managed IT complexity and avoids the problems associated with the native support of BYOD hardware. It also provides another opportunity for the IT department to become obsolete. Making sure an organization takes advantage of the former and isn't buried by the latter will differentiate the IT organizations that retain control of their computing clients from those that don't.
Those of us who have been around long enough have seen this attempted several times. In the past, the solution was either too software-centric, with no tuned hardware; or it was too hardware-centric, with thin clients running no software. As Goldilocks would say, for once this feels like it might be just right.