By: Paul Robson
Perhaps it feels as though the concept of cloud computing has been around for a long time, and that it’s an expected part of every CIO and CTO’s toolkit. It might be a surprise then to learn that cloud uptake is only at about 20 per cent, even though last year, spend grew by almost 50 per cent.
It could be argued that individual consumer adoption of cloud has been faster and more widespread than at enterprise level. Simply ticking the box on your phone that enables iCloud photo makes you an instant participant in the cloud revolution. Of about 1 billion iOS devices in market, there are about 782 million iCloud users. That’s significant uptake, especially when you consider it generates 200,000 iMessages per second.
Additionally, the services provided by companies such as Dropbox, Amazon and Netflix are only possible because of cloud technologies. People have become very used to the advantages of the cloud, so it shouldn’t come as a surprise when employees begin asking for the same conveniences from their workplace.
Initially, commercial cloud computing was simply a way to offload applications from a data centre to a more elastic compute-and-store environment, with the promise of greater performance on demand and/or lower costs.
Each iteration of the cloud has improved on that basic premise, adding features and reducing costs in line with Moore’s law (the number of transistors per square inch on integrated circuits has doubled every year since their invention).
Adopt early and innovate
This “lift-and-shift” trend has provided a glimpse of the benefits for organisations. If technology teams can save money on infrastructure, where can that money be invested to improve the customer or employee experience?
What intellectual capital is being freed up to think about what’s next, rather than the heavy lifting required to maintain the “boxes and wires”? And as system complexity and security concerns increase, where does organisational core competency end and smart outsourcing begin?
Increased accessibility of cloud infrastructure means start-ups as well as existing application providers have been able to move their offerings and innovate in the process.
However, for all the benefits, the data shows us cloud computing is still in its early days. The proportion of businesses adopting cloud as the primary compute-and-store environment is still in the minority, and this leads to some important concerns.
First, if cloud is offering such important benefits, how is your business taking advantage of these? Second, what are the likely opportunity costs for not moving to the cloud now, especially if your competitors or disruptors are already doing so?
Ready and enabled
Cloud computing has reached turning point. Financial services firm Morgan Stanley says cloud computing (and supporting technologies such as mobile phones), currently at 20 per cent of all workloads, is set for an explosion in growth.
The essential preparatory work of developing reliable, high-bandwidth internet (both wired and wireless) is well advanced in many Asian countries and compute-and-store in the cloud is cheap enough to indicate both stability of supply and elasticity of capacity. Security remains vitally important, but even intelligence agencies are using cloud tools – so for most, these concerns are being met.
Cloud computing is ready to step forward as the great enabler of all the key technologies in coming decades. It is time to shift focus from the cloud as “just plumbing” to a foundational, enabling principle of connectivism.
Connectivism is the network effect. The more nodes of any network that can exchange information, and the more rapid that information exchange is, the more efficiently that network can function. History is littered with examples: the Romans built straight roads for efficient movement of armies; railroads opened up the American west, until they were outpaced by the telegraph which transmitted information. The internet then displaced the telegraph as a super-efficient way to transfer highly complex ideas at near light speed.
Access not presence
Cloud computing is exciting because it abstracts the whole of a process rather than just its output: your server no longer lives in your data centre and presents just a web-page to your customers in the outside world; rather, the code for your web page, and the content-management system (CMS) it runs on, is location-less.
Where physical addresses denote a specific location, today an IP address is a point of access, not a point of presence. This distinction – where the physical becomes digital – presages what the cloud will enable.
It’s this notion of connectivism and abstraction that makes “server-less” computing possible. Service-oriented and micro-service architectures become viable when application programming interfaces (APIs) are made available at light speed and without location constraints. Connecting more and more nodes in the cloud gives rise to ever more possibilities. For example, Google Maps is a foundation (plumbing), but its real value lies in the real-time routing options from Waze (utility), powered by the crowd, in the cloud.
Machine learning gains ground
Machine learning is accelerating precisely because of the cloud’s capacity to house and process so much data so quickly. Training deep neural networks is a direct mimic of the “training” we have provided to the internet in the past few decades by way of our searches, image uploads and highly resilient architectures. These activities are repetitive and occur on a vast scale.
Using these very large volumes of data and very powerful computing power, we can discern patterns then test again for these patterns using new data. Patterns may sound neutral, but machine learning is already delivering amazing results in healthcare, smart cities and legal processes – all due to the commodification and scale of the cloud.
Productivity from the cloud will emerge from many directions. Cloud technologies are already freeing employees to work more efficiently, remotely, and with smart assistance.
Beyond the clear opportunities for customers and employees, it seems reasonable to forecast that the major technologies and the breakthroughs they will bring will all be dependent on the cloud. Autonomous vehicles require and produce data, and may need to communicate at light speed with other vehicles around them.
Virtual, augmented and mixed realties are cloud dependent, and, by nature of being multidimensional, could perhaps even be seen as a view into the cloud. Blockchain and Internet of Things (IoT) technologies are fundamentally cloud based. Perhaps the most interesting aspect will be the effects of these technologies in combination – for example, machine learning and healthcare for more accurate radiography.
The cloud is really just getting started. Its ability to enable other technologies of intelligence and scale are unprecedented, and already it’s possible to see how powerful and useful these may become. The challenge is how quickly your organisation is adopting the fundamentals so it can deploy the right tools and be part of the resulting acceleration.
Paul Robson is president at Adobe Asia Pacific.