A New Vision of the Cloud: Hyperscalers Moves in Hybrid and Multi Cloud Deployment for IoT

Organizations and industries across the globe are striving to digitalize their operations and enable their businesses to extract maximum value from the data. However, the digitalization and data collection efforts are now facing new roadblocks of accessing and managing vastly spread infrastructure. Enterprises and organizations that actively moved to the cloud haven’t changed the entire infrastructure, meaning that their functional architecture is scattered between cloud, on-premises, and edge computing infrastructure.

When it comes to the Internet of Things (IoT), ABI Research defines hybrid cloud services as the deployment of data and applications across multiple locations and cloud types (on-premises, public, and private). In contrast, multi-cloud services refer to enterprise use of services from different cloud service providers. Back in 2018, it was expected that the number of hybrid offerings would increase, and toolsets for enabling multi-cloud markets will grow. However, back then, the top cloud hyperscalers—Amazon Web Services (AWS) and Microsoft’s Azure—had close to no offerings for the hybrid cloud deployment. Now, both vendors have an end-to-end hybrid IoT deployment offering, such market movement was joined by the GCP. Interestingly, the hybrid cloud required the data to be as close to its source as possible to benefit from lower expectancy and the enhanced logical subdividing of data. Which means that cloud hybrid deployment for both private and public cloud offerings would create a large influx of investment and demand for edge cloud computing.

AWS introduced a hybrid-cloud partnership with VMware and provided an integrated cloud offering, jointly developed by the two companies, using Amazon Relational Database Service (Amazon RDS) on Vmware. For the IoT hybrid cloud offering, AWS offers a "cloud in a box", which the customer can deploy wherever it is needed. AWS announced its IoT Greengrass 2.0 will provide an open-source edge runtime, which includes a set of pre-built software components. AWS Outpost provides a consistent hybrid cloud solution that brings the same AWS infrastructure, services, Application Program Interfaces (APIs), management tools, support, and operating model. AWS’s perspective in multi and hybrid cloud are IT-focused—where there are so many IT components outside of the data center, devices, factories, and even cloud. Meaning they are actively offering edge-cloud synergy to enable flexibility and agility for the Informational Technology (IT)/Operational Technology (OT) infrastructure. At the same time, Azure has picked up a different positioning for multi-cloud spaces, where they intend to build a "telecom-grade cloud", though the claim is not to take or go into telecommunications business. In fact, Azure is interested in providing workload, edge processing, orchestration, and security services through its hybrid and multi-clouds, via its Azure Arc, Azure Stack, and Azure Stack Edge offerings. Additionally, Azure Arc is currently couples with Azure data services, such as Azure SQL Managed Instance and PostgreSQL Hyperscale

Interestingly, Google and its Google Cloud Platform (GCP) are gaining more traction for its analytics offering, although GCP haven't made a significant move in the IoT-specifically analytics in a couple of years. However, what is particularly interesting about GCP is that they recently revamped their Anthos offering. Google'sGoogle's Anthos is a Kubernetes-based suite of GCP cloud portfolio, which enables enterprises and organizations to manage workloads running on the third party (such as IBM, Azure, and AWS) as well as run applications on-premises and in Google'sGoogle's public cloud. This is potentially a disruptive move from GCP, as they are stepping up this game in hybrid/multi-cloud space. As everyone seems to be targeting their hybrid/multi-cloud offering for telecoms, GCP is no exception, as they have partnered up with AT&T and Spanish Telefonica to expand their global network and partnership for the edge marketplace. IBM, which was considered the leader with an all-in on hybrid cloud and AI, has been viewed at the reform of multi-cloud domain. But with Azure, AWS, and GCP entering the market, IBM is facing a significant challenge. Its partnership with Red Hat IBM Cloud with a technology base includes security and portability across multiple clouds and enables IBM to scale rapidly for its services.

The complexity and hype around multi and hybrid cloud deployment is that IT solutions have several essential factors for the use cases: proximity of the source and a need to be hosted on-premises for mission critical and regulated data. For example, an oil rig or wind farm with a highly unreliable network connection will require an IoT platform to run locally to avoid disruption due to connectivity issues. A private cloud on-premises deployment is often connected with services running on a public cloud for high-level analysis or planning. Additionally, IoT solutions can have a widespread geographic deployment in Europe, Asia, and the Americas, requiring local cloud providers for data privacy regulations. However, having enterprises be at the very different stages of digital transformation and migration to the cloud world, means the ability to deploy IoT platform services on existing service providers will be a deal-breaker. All and all, for the IoT Platform-as-a-Service (PaaS) vendors and enterprise's mid-way migration journey, the multi and hybrid cloud offering would be the next step for blurring the lines of vendor lock-in concept.