Get Ready for Infinite Computing

As cloud technologies evolve and proliferate, analyst Holger Mueller sees a future where limitless connectivity, data science and computing power let enterprises mine the vast potential of machine intelligence and deep learning.

By Tom Mangan

By Tom Mangan February 6, 2020

Holger Mueller isn’t afraid to say it: We’ve entered an age of infinite computing.

Mueller, vice president and principal analyst at Constellation Research, has made a career of understanding the impact of next-generation applications on enterprise IT. He understands people get skeptical at the notion that computing limits have vanished.  

“Now, any nitpicker would say, ‘nothing is infinite, ever’,” he said in an interview. “But my position is, when you can’t count it, and you can consume it as much as you need it for your business, it becomes practically infinite.”

By “it,” he means computing power, which is becoming easier to access. Mueller said this is forcing a major mindset shift for enterprise IT professionals.

He notes that conventional IT decisions worked something like this: Companies bought ERP systems to confront peak demands for closing books, manufacturing products and running payroll. Buying too much wasted precious corporate resources. Buying too little handcuffed system users.

Mueller said infinite computing helps IT people get off that tightrope.

“All of a sudden, you don't have to size anymore,” he said.

Size is irrelevant because the traditional constraints of IT — connectivity, bandwidth, storage and compute — have essentially disappeared. These forces also lay the groundwork for advanced automation and learning algorithms that can reinvent the enterprise IT ecosystem.

The five layers of infinite computing

In October 2019, Mueller published a research report titled “Infinite Platforms Power Enterprise Acceleration.” The report outlines how five technology layers enable infinite computing:

  • Connectivity – With more than 5 billion mobile accounts worldwide, most people who have the resources to connect to the internet can do it — and we’re closing in on connecting everybody else. “We can connect businesses with video and voice in a way that we couldn’t 10 or 20 years ago. So that's a given,” Mueller said in an interview. All these connections generate volumes of data that enable the next infinite computing layer.  

  • Questions and answers – Enterprises need business insights — monitoring and analyzing performance — to clear a path forward. In years past, reporting applications and data warehouses had their limits: Answers were confined to questions they were designed to handle. Seeking new questions and answers proved costly. Then Hadoop and the platforms it enabled came along, removing the old Q-and-A constraints and making data warehouses outmoded. As Mueller wrote in his report: “For the first time in mankind’s history, leaders can get all information into a single place, and then ask questions to obtain crucial insights — without knowing the questions beforehand.”

  • Compute  – Matching compute investments with demand has long been a frustrating exercise, Mueller noted in his report. Companies either bought too little and faced crushing slowdowns or bought too much and idled had three-quarters of their compute capacity. Today, cloud technologies, virtualization and hypervisor software allow companies to spin up only the resources they need without investing in excess capacity. That brings CPU costs down dramatically for enterprises while motivating cloud-services companies to innovate in ways that maximize capacity usage.  

  • Machine learning – Infinite connectivity, Q&A and compute pave the way for infinite machine learning. Constraints such as stale data and limited CPU availability that held up machine learning in the past are gone, Mueller wrote in his report. This shift allows enterprises to develop next-generation applications that take full advantage of the most advanced algorithms. That, in turn, gives companies access to the most transformative technologies.  

  • Deep learning – In the ultimate layer of infinite computing, machine learning algorithms become sophisticated enough to evolve without human intervention. “The step, while small on the technology side, has a massive impact on the business side,” Mueller wrote. Algorithms can teach IT systems to adapt to ever-changing marketplace environments, while IT leaders can shift more resources to the cloud and save money by transitioning to an operational-expenses mode from a capital-expenses mode. Recruiting scarce IT and data talent is less of a challenge, Mueller wrote, because deep-learning systems can figure out how to automate many tasks.

What Infinite Computing Means for Businesses

Infinite computing doesn’t remove all technology constraints. Protecting sensitive data often will require on-premises data storage. Renting public cloud services may cost more than buying server stacks in some scenarios. Indeed, the wide-open future of infinite computing may have some IT pros longing for the limitations of simpler times.

Nevertheless, infinite computing is a force to be reckoned with because it can deliver solid enterprise value. Moreover, these next-generation technologies are helping companies reinvent their business models.

“They can't be really measured in the traditional, I jokingly say, ‘bean-counter’ ROI models,” Mueller said. “It's all really about changing your company in different directions — where you think best practices can be enabled and new markets can be created.”

At a more granular level, infinite computing expands the flexibility and agility of enterprise IT departments. Companies can limit vendor lock-in and bring portability to their IT workloads.

“What's being provided right now by vendors, like, for instance, Nutanix, is the portability across different public cloud permutations and on-premise private cloud permutations,” Mueller said.

The ability to select and implement cloud technologies pragmatically to meet a particular company’s needs is critical, according to Mueller. Getting locked into one or vendor can hinder a company’s adoption of hybrid cloud. 

He sees building hybrid clouds as crucial because it enables — and encourages — workload portability.

Portability, in turn, can help companies avoid getting stuck with vendors that fall behind in the relentless scrum for cloud dominance.

“You don't know, actually, who's going to be the winner in the cloud game,” Mueller said. “You don't want to be tied into a losing platform.”

Companies have to find the right deployment platforms and build code that’s transportable across those platforms.

“Workload portability is very, very important to CIOs and CTOs today — so you don't have the lock-in effect,” Mueller said.

As infinite computing culminates in deep learning, IT leaders will be able to respond rapidly to change.

“Artificial intelligence changes everything in the enterprise to levels where we can't think what could be computed right now,” he said.

Automated learning algorithms can overcome natural human biases. Deep learning systems can plug the talent gaps for companies that lack access to the best minds in computing and data science.

As Mueller concluded in his report on accelerating enterprise technologies:“The biggest benefit is closing the insight-to-action loop, taking slow reaction times, human delays, processing errors and oversight out of the equation of business success.”

Tom Mangan is a contributing writer. He is a veteran B2B technology writer and editor, specializing in cloud computing and digital transformation. Contact him on his website or LinkedIn.

© 2020 Nutanix, Inc. All rights reserved.  For additional legal information, please go here.