It’s Time To Distribute The Cloud
Who said you can’t have your cake and eat it too? Organisations need not restrict themselves to one iteration of IT infrastructure, as distributed cloud is offering the best of every world.
An Untapped Behemoth
From Turing’s mechanical beasts to the very first microprocessor, IT has seen a number of fundamental changes. And while every revolution and uprooting of a previous model accelerated our abilities, each innovation also brought new challenges. On premise computing, for example, began to prove costly and inefficient due to the large amount of downtime that occurred. This was one of the key reasons the cloud saw such growth, as it provided a more flexible and cost efficient solution.
Undoubtedly, cloud infrastructure is here for good. In fact, Gartner predicts that public cloud spending will reach $331 billion in 2021, with this demand only growing further. Why has the investment still not stopped? Well to answer that let us look at another statistic. A recent study by McKinsey estimated the value of business improvement offered by the cloud to be more than $1 trillion across Fortune 500 companies by 2030. What this tells us, is that we’re barely scratching the surface in regard to the benefits of the cloud.
Rising Costs and Technical Limitations
As the cloud started taking over virtually every IT system, its novel nature required attention. Change rarely happens smoothly and with this major overhaul, imperfections were rife. Initially, one of the key attractions of the cloud was its flexibility in cost compared to local solutions, as organisations could pay as and when they used services. However, new research is showing that while the cloud offers improvements earlier in a company’s life cycle, as they grow and mature, the costs begin to outweigh the benefits in some areas. Because this change only becomes visible later on in the timeline, restructuring can seem overbearing.
While there’s been a back and forth between when and where to use the cloud for its financial advantages, one thing is clear: this is not an all or nothing decision. Different companies and different lifecycle stages will require varying combinations of cloud, hybrid and on-prem solutions. Cost, however, is not the only issue to be ironed out. Emerging technologies are pushing the limits of the cloud and they are demanding a refinement of the model for their continued development. Building and running on the cloud using things like containers is proving exceedling complex for these use cases.
Real-time Big Data Streaming, IoT and AI/Machine Learning
Each of these phenomena are gigantic industries in their own right, but ultimately they all rely on data connectivity. Take IoT for example, where connected devices may communicate large amounts of data to construct a digital twin of a physical system. Often time sensitive properties demand low-latency for them to be accurately represented, which current cloud platforms can struggle to provide. Likewise, training AI and ML almost always consists of big data sets, which communicating through remote data centers can prove difficult.
Simply put, even in a hybrid version, our cloud needs to change to support these examples. Scaling, latency and setup/configuration are all experiencing issues when running these kinds of applications on purely cloud environments.
Enter Distributed Cloud
The significance of these technologies means that enabling the cloud to sufficiently support them is essential. In a sentence, distributed cloud gives an organisation the option to enable any public or private cloud, hybrid, on-prem or edge environment for its IT functions. Each part of your infrastructure can be mapped to the one it suits best, rather than having to choose a lowest common denominator. The edge, for example, offers a crucial solution to latency problems with the cloud by processing data closer to its source. Or consider the security issues around the cloud. With stricter GDPR regulations, storing sensitive data on public clouds became a concern and often using on-prem is preferable to maintain protection.
Distributed cloud is offering greater access to using these different solutions and, in many ways, could be the last word on the cloud, as future iterations can simply be adopted into the distributed model. But once again, a new model means new challenges. It requires a high degree of orchestration, and the lack of options regarding tools like containers is restricting its use. Compute platforms will need to simplify how we build applications for distributed cloud. Only this way can we access the huge increases in efficiency that the cloud offers.