Could anyone have predicted Uber would lose its founding CEO or the price of bitcoin would jump 20fold over the course of 2017? No doubt 2018 will be just as unpredictable, as a huge range of tech trends from the cloud and blockchain to machine learning and virtual reality combine and collide. This is the first in a series of predictions by SiliconANGLE’s staff, supplemented by other experts, on what’s coming next year in enterprise and emerging technologies and the broader tech industry.

One thing you can say about the cloud infrastructure market is that it never gets boring. Just as enterprises became comfortable with moving large-scale workloads from their data centers to the cloud, along came new options to go the other way. To public, private and hybrid cloud you can now add the “multicloud.” Meanwhile, the market is also going vertical in a big way. Here are four trends to watch for in 2018:

Multicloud mayhem

International Data Corp. estimates that more than 85 percent of enterprise information technology organizations will commit to multicloud architectures by 2018, spurred on by the promise of flexibility and cost advantages that accrue from shifting workloads seamlessly between platforms. However, AtScale Inc. Chief Executive Officer David Mariani labels the multicloud dream a “fallacy.” Cloud infrastructure providers have no interest in making it easy for their customers to move off of their platform, he asserts, so they will do everything they can to lock them in.

I tend to agree. If you look at the volume of announcements that came out of Amazon Web Services Inc.’s recent re:Invent conference, you see a vendor doing everything it can to make its services so compelling that customers won’t want to go anywhere else. Amazon will use shock and awe to make its cloud the biggest and most diverse. Microsoft Corp. will leverage its strength in applications, and Google LLC will appeal to data scientists and those seeking the lowest cost. Oracle Corp. will have a database-centric story.

But the infrastructure companies don’t necessarily hold all the cards. Suppliers such as IBM Corp. and VMware Inc., which will never challenge the big three incumbents, will set themselves up as impartial multicloud brokers, while startups such as CloudVelox Inc. and open source advocates including Red Hat Inc. and Pivotal Inc. will promote their platforms as foundations for multicloud bliss. Containers also promise to ease the transition to multicloud by separating the application from the underlying infrastructure. It promises to be a free-for-all for a while, with customers enjoying a wealth of options, but also hard choices.

How others see it

• Dell Technologies Inc. agrees that a certain amount of chaos is inevitable. “As more applications and workloads move into various clouds, the proliferation of cloud silos will become an inevitability, thus inhibiting the organization’s ability to fully exploit data analytics and AI initiatives,” said a missive from the company’s public relations agency. “This may also result in applications and data landing in the wrong cloud leading to poor outcomes.” Dell sees the need for what it calls a “mega cloud,” which offers “a federated, intelligent view of an entire IT environment,” to enable IT organizations to easily shift workloads around. You can see how, not surprisingly, that plays nicely to its product line.

• CloudVelox CEO Raj Dhingra sees a need for his company’s brand of automation to handle multicloud complexity and for enterprises to seek out managed service providers and hosting partners such as CenturyLink Inc. and IBM to manage the whole mess “because they cannot solve things internally.”

• “Fear of cloud lock-in will result in cloud sprawl,” posits Satyen Sangani, CEO of Alation Inc. “Finding information or code within a single cloud is tough enough. The data silos built from multiple clouds will be deep and far apart, pushing the cost of management onto humans.”

• “Multicloud becomes a mess if you keep your focus at the virtual machine layer,” said Vince Lubsey, senior vice president of innovation at Virtustream Inc. “If you look at it from a container layer, then it becomes easier to move your workloads. It’s a matter of services and access to data.”

On-premises and public cloud distinctions will blur

Azure Stack was Microsoft’s brilliant answer to the compatibility fears customers have about moving to the cloud. It proposes to put a duplicate of its own cloud on customer sites and let them move at their own pace. AWS and VMware are doing effectively the same thing with their joint migration initiative. You might call both a form of vendor lock-in, but for customers that are comfortable riding with a single vendor, these initiatives are great training wheels. Expect all three of those companies to grease the skids in 2018, and for Google to make its strategy known. In the open-source world, creative packaging of options such as OpenStack and Cloud Foundry will target the same needs.

How others see it

• “Extending on-premises computing to the cloud has been a bit messy,” said Virtustream’s Lubsey. He thinks a migration strategy that starts in the cloud has a better chance of success, particularly when incorporating distributed internet of things devices. “We think we’re going to see more hybrid offerings like Azure Stack in the future,” he said.

The cloud will go vertical

AWS’ Sumerian service, IBM’s Quantum Experience and Google’s Cloud Video Intelligence are three examples of how cloud infrastructure services are becoming more specialized. Expect to see many more. Established cloud vendors will use vertical offerings to entice new customers, while second-tier providers and newcomers will look to carve out new markets. High-performance computing will be a particularly hot area, since the cost of building out those capabilities on-premises is so high. So will anything related to artificial intelligence and machine learning. This trend is a windfall to small companies working in complex areas such as genomics and fluid dynamics, since they can get affordable access to technology that was once available only to the largest corporations.

How others see it

• Cloud infrastructure companies have always coveted developers first and foremost, and the rise of vertical clouds will only intensify those efforts, said Peter Cho, vice president of product at Arigato Machine Inc., which does business as Manifold. “Cloud services will fail or succeed based upon one factor: developer experience,” he said.

The ‘internet of things’ will redefine the cloud

Cisco Systems Inc. calls it “fog computing.” Others call it “edge computing.” Whatever the terminology, the need to accommodate sensors and intelligent devices at the corners of the network will change the way organizations approach the cloud. The inherent latency and delay of the centralized model won’t lend itself to the split-second decision-making needed, for example, to decide whether to take a failing telecom switch in the field offline. New cloud architectures will accommodate localized processing with a subset of data sent to the host. Amazon Greengrass is a first stab by the cloud giant at this evolving need. The rapid growth of IOT devices will quickly attract others, and noncloud companies such as sensor and database makers will also get in on the action.

How others see it

• “Rather than moving the data, the applications and resources needed to process it will be moved to the data,” said Mark Bregman, chief technology officer at NetApp Inc. “This will have significant implications for new architectures like edge, core and cloud. The amount of data ingested in the core will continuously become less than the amount generated at the edge.”
• “Traditional industries like manufacturing, which have been struggling for differentiation, will embrace edge analytics to drive new revenue streams and/or significant yield improvements for their customers,” said Sastry Malladi, chief technology officer at FogHorn Systems,

Don’t expect major disruptions to the overall cloud landscape anytime soon. Forrester Research Inc. expects AWS, Microsoft and Google to own 80 percent of the platform market by 2020. However, new niches will continue to emerge, and each has the potential to give birth to its own leaders – and potential disrupters.

Image: Flickr CC