Why the supercloud must become a reality

The IT environments of today can be highly complex. Organizations often deploy multiple cloud services and still run some applications on-prem. Companies are eager to make the management of all these workloads smoother and cheaper. The concept of a “supercloud,” a layer of governance that sits above all IT operations, sounds appealing at face value. What exactly does it entail? And also: how do you approach it in practice as an organization?

On paper, an IT strategy around hybrid multi-cloud sounds most logical. Companies want flexibility and hope to be able to choose the best solution for each individual workload. Sometimes this means on-prem is most convenient or an application performs best within AWS, Google Cloud or Azure. However, due in part to the multiplicity of choices, consistently managing multiple cloud environments and on-prem is difficult in practice. In addition, companies aim to achieve process optimization wherever possible. What if you could automatically choose the right environment for each application and thereby drive your IT environment much more efficiently?

Nothing happens on its own

Moving applications and workloads isn’t an automatic process. IT vendor Equinix cites that migrating from one cloud to another is usually expensive and difficult. In essence, companies are getting penalized for making the wrong choices because they can’t correct them quickly. As a result, hybrid multi-cloud in practice falls far short of the potential it theoretically possesses. That’s why Equinix brings up the concept of the supercloud: this would make it possible to place cloud workloads anywhere, with the help of an “abstraction service” that is coded once and then can be deployed anywhere. This would have to be compatible with any cloud environment and automatically and seamlessly hook into it. Currently, this is not yet a reality, partly because the exact manifestation of the supercloud is not yet known.

At Equinix, they refer to a “cloud Switzerland”: a neutral party that can pass on the functionalities of all cloud services to the end user without any friction. Customers benefit from this because it circumvents vendor lock-in from a single cloud service. In addition, they would not be dependent on the availability and costs that apply to the chosen hyperscaler. To make that desirable state of affairs possible, however, it is necessary to maintain visibility and control over corporate data.

Data is the key

Enterprise data is enormously valuable and can provide many new insights, often aided by AI. However, organizations are wary of losing control of this data. Parties such as VMware and Nvidia are responding to this with Private AI Foundation and Google with compliance initiatives around Vertex AI. However, more is needed to make the supercloud a reality: the data should not only be available on-prem or in a particular cloud, it should be able to be safely guided from cloud to cloud.

A party like Snowflake provides an important building block for the supercloud in this regard. This company’s data platform is “cloud agnostic” and thus works with Google Cloud, AWS and Azure. Thus, it has essentially already created the abstraction layer to deploy data wherever consumers want it. Data governance can therefore be regulated separately from hyperscalers thanks to Snowflake. The cloud services in question are still all “walled gardens” that cannot consistently communicate with each other, but the bridging is still getting done – albeit indirectly thanks to the likes of Snowflake and others.

The nature of the supercloud

According to Dave Vellante of theCUBE, Snowflake is “the most prominent example of a supercloud.” He also characterizes the initiatives of Databricks and VMware as major trailblazers of the concept. In other words, we are already seeing building blocks of the supercloud emerging, although the parties are not yet working together to solve certain compatibility issues. In fact, in his pieces, Vellante talks about supercloud as the next step after multi-cloud, where interconnectivity provides additional value on top of separate applications and cloud environments. In a sense, he says, the supercloud is already here, albeit to a limited extent.

In fact, the supercloud is a very interpretable concept. It can be understood as a concept that contains gradations that applications may or may not meet. Can a solution connect seamlessly to any cloud and remove this complexity for the user? Then it delivers certain supercloud principles. However, a player that can connect all the dots is still missing. Achieving that would depend on a broader understanding of the supercloud as the aforementioned cloud-Switserland: a true umbrella piece of software that makes it easier for companies to connect their own IT environment to cloud services.

Organization can currently achieve only certain supercloud aims. To make this happen, they choose solutions with open standards and a high degree of compatibility, and this offers them the chance to simplify IT management and remain flexible with their choices of vendors. Yet the promise of the supercloud is bigger than that and has not yet been fulfilled. In other words, there are still opportunities there for innovative companies.