While innovations of the web-scale consortium of Facebook, Amazon, Netflix, and Google (FANG) are impressive, it’s challenging to take those innovations and apply them to enterprise IT.
For example, take Facebook’s strong contributions to the Open Compute initiative. Facebook’s most recent contribution is a 100Gbps data center switch and, while the switch may have limited applicability to most regular enterprises, it does highlight the growing value of IT infrastructure-specific intellectual property created by end-user operations.
Digital transformation drivers
As I’ve talked to end-user organizations, a significant consideration is the speed at which businesses are adopting technology to solve specific business challenges, and the ability of internal IT to keep pace. Facebook and other web-scale providers face similar problems.
Traditional enterprise network gear is not designed to scale to the levels required by the FANG consortium. In a video, AWS highlighted the challenge of creating a logically segmented network using traditional network products. AWS came to the conclusion that they were beta testing commercial equipment for most of their use cases.
Facebook recognized early on that the development cycles for traditional networking gear would not provide the capacity needed to meet their growth demand in users and services. So, the company launched a team of engineers focused on building data center components specific to their use case.
The 100Gbps data center network switch is a result of the inability of traditional providers to keep pace with the demand of Facebook’s business.
There are very few enterprises that have the data center scale or data requirements that necessitate 100Gbps data center switches. Many enterprises are slowly rolling out 10Gpbs data center connectivity. A consistent challenge is the ability for end-user IT infrastructure to keep demand with the speed at which business requires innovation.
A constant frustration within enterprise IT is the process in which lines of business (LoBs) procure IT resources directly due to the inability of enterprise IT to deliver a frictionless experience in building new applications. Further discussions I’ve had with both large and small organizations is that many of these business-driven IT initiatives fail to achieve the desired results.
Unavoidably, LoBs discover IT operations are complicated. Public cloud, for example, doesn’t eliminate the need for security, patch management, or audit responsibility. LoBs find themselves returning to IT to solve these traditional challenges in the LoB’s IT initiative.
IT organizations turn to enterprise vendors to help with the speed of innovation needed to support the business initiatives. In my discussions, IT organizations are finding a similar problem as the FANG group. Enterprise IT vendors are not positioned to innovate at the pace or the granularity of specific industry or organizational requirements.
Out of necessity, IT organizations are beginning to embrace open source solutions to meet the challenge. Open source projects tend to move at the pace required by the project contributors. The contributors are starting to look much different than many traditional open source projects. Many forward-thinking end-user enterprises are dedicating developer resources to contribute to these projects directly, or take the base solution of a project and add capability.
Whether it’s direct contribution or internal development, IT infrastructure is discovering the need to directly contribute to the low-level innovation necessary to meet the business requirements.