5 Trends giving cloud a boost
State of Cloud: Trends & Predictions
Most companies have already shifted some processes to the cloud to reduce costs and speed up service delivery. Recent developments in several key areas are putting wider adoption and additional rewards within closer reach.
A blog post by David Schatsky, managing director, Deloitte LLP; Amit Chaudhary, principal and US cloud engineering lead, Deloitte Consulting LLP; and Amanpreet Arora, manager, Deloitte SVCS India Pvt. Ltd.
Ninety percent of enterprises have adopted the cloud to some degree, according to the Deloitte US CIO Program’s 2018 survey of global CIOs, but only about 20 percent of business processes are currently running there. That dichotomy suggests many companies have a ways to go in using the cloud for purposes beyond lowering IT capital expenditures and accelerating service delivery—purposes such as innovation and transformation.
Five vectors of progress currently at work, however, could increase cloud adoption and help companies derive additional value from the technology.
Many enterprises are rapidly stepping up their spending on AI: About half of respondents to a recent Deloitte survey expect to increase their AI investments by 10 percent or more in the next fiscal year.
Many cloud vendors are stoking demand for AI technology by offering tools and services that make it easier to develop, test, enhance, and operate AI systems without big upfront investments. These include hardware optimized for machine learning, application programming interfaces that automate speech recognition and text analysis, automated machine learning modeling tools, and AI development workflow platforms. Shipbuilder Samsung Heavy Industries, for instance, is using Amazon Web Services capabilities including machine learning to develop and manage autonomous cargo vessels.
The symbiosis between the cloud and AI is accelerating the adoption of both: By 2025, AI is expected to account for as much as 50 percent of total public cloud services revenue.
Automated Data Classification
Traditional IT environments employ a perimeter-based approach to security, where the aim is to keep out bad actors. The cloud approach, on the other hand, relies on authentication and authorization, with a focus on identifying who can do what with which assets. Because of that difference, companies that simply “lift and shift” legacy workloads to a public cloud can face unwanted data exposure and regulatory noncompliance.
Automated data classification tools can help by categorizing workloads and data to be migrated based on sensitivity and business impact. Many major cloud providers offer classification capabilities that integrate with their solutions; security providers have also partnered with cloud providers to deliver them. Increasingly, these tools use machine learning to more effectively classify content such as personally identifiable information or sensitive commercial, military, and government data, expanding the range of data and apps that can be migrated to the cloud with confidence. For example, automated data classification tools allowed Edmunds, a car shopping and information platform, to better discover, classify, and protect its data in the cloud; administrators received actionable information about potential threats, including those that would ordinarily have been difficult to discern.
All of this can enable a more secure migration of legacy systems to the cloud as well as improve compliance with regulatory standards such as the Health Insurance Portability and Accountability Act and the General Data Protection Regulation. Globally, spending on data classification solutions is expected to grow by 25 percent annually.
Some companies may not want to confront the challenges of migrating core legacy applications customized over the years to meet industry-specific requirements; others may want to upgrade their systems to support modern operating models without necessarily building everything from scratch. In either case, industry clouds can provide an attractive pathway for taking advantage of the cloud.
Industry clouds are becoming increasingly popular, leading many cloud vendors to expand their portfolio of offerings. In addition to providing a ready-to-use environment with tools and services tailored to a specific sector’s operational requirements, these clouds can often aggregate data from multiple clients, potentially enabling vendors to offer additional benefits such as industry insights or benchmarks. More than half of respondents to a recent survey (64 percent) either already use industry-specific cloud services or plan to do so in the next 12 months.
DevOps and Beyond
IT organizations frequently look to the cloud to reduce costs and increase business agility, but traditional IT operating styles—manual and siloed processes to deliver monolithic applications—can limit these benefits. New IT operating models such as DevOps and paradigms such as cloud-native application development are helping organizations wring more value from the cloud.
DevOps has gained considerable ground in recent years for its ability to help IT organizations deliver technology faster and more frequently with better quality and stability. For some enterprises, IT operating models are evolving further into NoOps and serverless environments in which cloud providers automate many systems administration tasks and take over infrastructure and security management. This can yield substantial cost savings and other benefits.
Meanwhile, cloud-native development leveraging containers and microservices, though still nascent, is also on the rise. In 2018, 20 percent of IT decision-makers were focusing primarily on new cloud-native development, according to one survey, up five percentage points from 2017. Education services provider Kaplan’s adoption of microservices and containers for its applications in the cloud reduced the number of virtual machine instances by 70 percent, resulting in 40 percent cost savings per application. Containers also enabled automation of application update deployment, reducing the time required from hours to minutes.
Finally, migration to a public cloud can be complex, time-consuming, and expensive because it may necessitate substantial changes to legacy apps. For many companies, not only are those legacy apps not ready for the cloud now, but they may never be ready, for reasons including architectural incompatibility, data ownership, and compliance. Enter hybrid cloud architectures, which can provide enterprises a way to take advantage of the cloud where appropriate while leaving some legacy applications in place on-premise.
Hybrid clouds are a mix of private and public cloud environments with a degree of workload portability, integration, orchestration, and unified management that allows workloads to be moved between the two environments. Using a hybrid cloud, enterprises can gain the benefits of the public cloud for scalability, business continuity, external collaboration, and access to cutting-edge technologies for innovation while retaining the on-premise option for certain legacy apps. A hybrid cloud infrastructure has enabled Deutsche Bank to run more than 40 percent of its workload on five percent of its total infrastructure, for instance, reducing the time needed to go from proof of concept to production from months to weeks.
The evolution of cloud computing has been remarkable, resulting not only in a rapidly growing market and vibrant innovation but also in shifting mindsets and practices pointing toward a new IT operating model and software development paradigm. These five vectors of progress suggest the evolution has only just begun.
This article first appeared on the WSJ