<< back Page 3 of 3

New Technologies Shaping Today’s Big Data World


Why it’s hot: The ongoing separation of compute and storage “is a game-changer that will drive innovation in data analytics over the next several years,” said Greg Betz, senior vice president of data intelligence and automation at NTT DATA Services. “It enables organizations to collect and store more data than ever before because storage costs are relatively cheap when separated from compute—particularly if the organization is using public cloud services. This allows organizations to collect and store both business data and infrastructure data indefinitely.” Underpinning this technology advancement are “greater network bandwidths with low latencies and the shift to cloud services which enables businesses of any size to leap into analytics without incurring the upfront costs of purchasing hardware,” he added.

Emerging or widespread? The adoption of data solutions which separate compute and storage “largely depends on the perceived business value of the technology and an organization’s investment in legacy systems where compute and storage are tightly coupled,” said Betz. “Small organizations with minimal investment in legacy data systems will very quickly adopt these solutions as their data volumes grow—which typically happen at an exponential scale—and storage costs rise. Large organizations that treat data as an asset and have started their digital transformation are typically the first ones to adopt this technology as they realize that a key aspect to unlocking data value is separating compute from storage.” The current push to working from home “will further accelerate the adoption of these technologies as many organizations look to public cloud solutions to maintain business continuity.”

Gotchas: Getting to the cloud to realize the benefits. Organizations that are not willing to use public cloud platforms “will not be able to realize the full benefits of a data system that separates storage and com-pute,” Betz explained. These benefits include cost savings for storage and the ability to almost infinitely scale the compute. “Data sovereignty rules are still an obstacle for many organizations. However, the push to work from home coupled with encrypting data at rest and in motion is lowering the barrier to entry,” Betz noted. 

In 5 years: “Separating data from compute will be the norm for any organization with large data volumes, which will be most organizations,” Betz said. “As data volumes grow, so too will the need to manage and govern it, so tools focused on cataloging, auditing, and tracking data will gain popularity or be baked into the system.” Such technology “will enable data-driven transformations, which in turn will drive further adoption and investment in this technology and lead to further transformations.”


Why it’s hot: “Many of the organizations that decided to skip the Hadoop bandwagon are either seriously considering, or have actually started going down the path of migrating their data to the cloud,” said Jayaprakash Nair, head of analytics—data science, machine learning, AI and visualization at Altimetrik. “The new applications being built in these organizations are being architected in a cloud-native manner.” Companies are looking for a more nimble way to get insights from data, and they see the managed data services landscape on the cloud as a key approach.

Emerging or widespread? Managed data services are “definitely not in the experimental stages—at the same time, they have not achieved its full potential yet,” said Nair.

Gotchas: Legacy applications. “Most of these applications were architected for on-prem deployments and the data that they produce also resides on-prem,” said Nair. “Now, if the organization decides to move the bulk of its data to the cloud, there needs to be a cluster of connections feeding the data from on-prem to the cloud. While this has already been successfully implemented by many organizations with good hybrid-cloud support from public cloud providers, the magnitude of this problem across the industry cannot be overstated.”

In 5 years: “We will see many organizations going from ‘cloud also’ to ‘cloud only,’” Nair predicted. “The price competition, as well as the economies of scale achieved by public cloud operators, will ensure that the overall price advantage of moving to the cloud becomes a much more compelling proposition. Many organizations will opt for a multi-cloud architecture.”

Looking Ahead

The last 6 months have accelerated the journey toward digital transformation for companies of all types, making it  imperative for them to expand their use of technologies that drive speed and efficiency and lower costs. Expect to see more reliance on processes and solutions that leverage automation of previously manual processes and improve time-to-insights, as well as more dependence on fit-for-purpose multi-cloud architectures.

Photo by Luca Bravo on Unsplash

<< back Page 3 of 3


Subscribe to Big Data Quarterly E-Edition