Advanced Computing in the Age of AI | Thursday, March 28, 2024

Predictions 2017: Advanced Scale Trends to Watch 

(Nikki Zalewski/Shutterstock)

“The future ain’t what it used to be,” so say the blasé and the pessimistic. But in advanced scale computing expectations are so charged that it’s reasonable to say that for our sector of the technology industry these are the good old days. There’s a sense we’re on the cusp of a Great Leap Forward in the enterprise adoption and adaptation of HPC technologies that will rival the major transitional points (the PC revolution, client-server computing, the Internet, etc.) of the past 40 years, one that builds on these achievements and delivers to businesses and consumers amazing new capabilities.

Now that it’s the first week of the New Year it’s time to speculate on major developments to be anticipated for 2017. We put out a general call for predictions and received a number of interesting responses. Even if some of these don’t come literally true, they doubtlessly point to key trends regarding the infiltration of HPC technologies into the enterprise that we need to watch in the upcoming year.

IoT: Wait til This Year

IoT has been a futures story since way back. But at least of one our futurists tabbed 2016 as the year it went mainstream (see below), and 2017 will see IoT and IIoT proliferate across a wide range of commercial and consumer markets.

While interoperability of devices remains a major hurdle, significant progress has been made on other IoT fronts, including at-scale data capture of machine data from sensors, the storage of multi-format data and most significantly, according to analytics and data visualization vendor Tableau, HPDA (high performance data analytics) on all that variegated data.

“With large quantities of IoT data now easily ingested into cloud storage, focus is shifting from capture to analysis,” Tableau reported in a recent white paper. “Organizations are demanding analytical tools that seamlessly connect to and integrate diverse forms of cloud-hosted data. IoT data tends to be heterogeneous and stored across multiple systems, from Hadoop clusters to noSQL databases. It’s no small feat to access and understand all that data. As a result, people need analytical tools that seamlessly connect to and combine a wide variety of data sources… into a single view…”

Amit Pandey, CEO of software-defined application services company Avi Networks, sees 2017 as a milestone year for IoT, stating that “IoT devices will go mainstream and mobile will set new records, driving greater traffic. Screen-free devices like Amazon Echo and Google Home will proliferate next year, thrilling consumers but also introducing new security threats and potential performance issues from increased application traffic.”

Also noting the risk/reward nature of IoT proliferation is David Galton-Fenzi, CEO of Zycko, an advanced networking specialist. He cited studies forecasting IoT revenues could exceed $470 billion for IoT hardware and software vendors (according to Bain Capital) and a doubling of installed devices to 31 billion by 2020 (according to HIS). But IoT-related security concerns him.

“The increase in devices and the amount of data that they send will place pressure on networks, but more disturbingly, there will be obvious issues around securing both data and devices. In the wake of the recent distributed denial of service (DDoS) attack on popular websites, such as Twitter and Spotify, using internet-connected home devices, network security will be high on the agenda in 2017.

Chetan Sharma, CEO of Chetan Sharma Consulting, said 2016 was the chasm crossing year for IoT, but challenges remain.

“IoT emerged as a big category last year and we are just starting,” he said. “While the potential is unlimited, industry needs to come together on a number of issues around fragmentation, security, policy, and privacy. The fact that we are selling more non-phone devices than phone devices on cellular networks in the US should be a hint of things to come. 2015-16 saw a wave of consolidation and we will continue to see stronger players get more aggressive this year across domains, which is very exciting.”

Big Data: Draining the Swamp

The evolution of data management in 2017 and beyond is also on the mind of John Schroeder, executive chairman and founder of Apache Hadoop-based software vendor MapR.

“In 2017, organizations will shift from the ‘build it and they will come’ data lake approach to a business-driven data approach,” he said. “Use case orientation drives the combination of analytics and operations.”

Schroeder described the typical data lake vision as: “Imagine what your business could do if all your data were collected in one centralized, secure, fully-governed place that any department can access anytime, anywhere.” Sounds attractive, right? Sure, but it “too frequently results in a data swamp that looks like a data warehouse rebuild” that can’t address real-time and operational use case requirements.

He cited the real-time needs at the individual level in several industries. Ecommerce sites must provide individualized recommendations and price checks; healthcare organizations must process valid claims and block fraudulent claims by combining analytics with operational systems; media companies are now personalizing content served through set top boxes; and auto manufacturers and ride sharing companies are interoperating at scale with cars and the drivers.

“Delivering these use cases requires an agile platform that can provide both analytical and operational processing to increase value from additional use cases that span from back office analytics to front office operations,” Schroeder said. “In 2017, organizations will push aggressively beyond an ‘asking questions’ approach and architect to drive initial and long term business value.”

Related to this are challenges around data governance and competitive advantage.

“In 2017, the governance vs. data value tug of war will be front and center,” Schroeder said. “Enterprises have a wealth of information about their customers and partners… Organizations are now facing an escalating tug-of-war between governance required for compliance, and the use of data to provide business value and implement security to avoid damaging data leaks and breeches.”

Processing all that data in (near) real time is the objective of much technological development. It’s why 2017 will be “The Year of the (D)RAM,” according to Tom Lyon, chief scientist at scale-out IT infrastructure builder DriveScale.

This year, he said, “will be the year that servers with 1TB of DRAM become commonly available – both in enterprises and in the cloud. Many datasets that were ‘Big Data’ five years ago will be handled easily in memory. But the truly large ‘Big Data’ datasets will still be growing faster than any improvement rates in hardware.”

Mesh Networks: Seamless and Secure

One of the most interesting futures comments came from Galton-Fenzi regarding “mesh network” architecture, an emerging technology that he said offers a more secure, stable network connection than the current internet network architecture. “They really do demonstrate the Holy Grail that is ‘seamless connectivity,’” he said.

“In a mesh architecture, distributed nodes ‘talk’ to each other,” he said, “and compared to today’s internet — which is based on a few centralized access points or internet service providers — the only way to shut down a mesh network is to shut down every single node in the network. This means there’s no single point of failure. The design affords a more robust, predictable network, with service providers better able to establish control.”

In addition, he said, mesh networks allow power to be more evenly distributed, increasing network redundancy. “The argument continues to grow for mesh networks,” he said, “particularly in cases where internet connectivity is threatened by natural or man-made disaster.”

Blockchain Breakout

Another technology whose year in the sun is predicted is blockchain, to date used mainly in financial services but ready, according to Juniper Networks, for other sectors.

“We expect that in 2017 a raft of proof of concepts will be developed to integrate the technology into a much wider array of applications, with logistics and identity management to the fore,” reported Juniper. “We also envisage that a number of national governments will instigate trials incorporating blockchain technology in a bid to automate manual processes which are time-consuming and expensive.”

The key to blockchain, said MapR’s Schroeder, is the trusted third-part “be-all end-all technology protocol” in the middle of all transactions, the brainchild of Nick Szabo nearly 20 years ago.

“This trust protocol provides a global distributed ledger that changes the way data is stored and transactions are processed,” Schroeder said. “The blockchain runs on computers distributed worldwide where the chains can be viewed by anyone. Transactions are stored in blocks where each block refers to the preceding block, blocks are timestamped, storing the data in a form that cannot be altered. Hackers find it impossible to hack the blockchain since the world has a view of the entire blockchain.”

Blockchain provides obvious efficiency for consumers, who won't have to wait for a SWIFT transaction or worry about a central data center leak. For enterprises, blockchain presents a cost savings and opportunity for competitive advantage. “In 2017, there will be select, transformational use cases in financial services that emerge with broad implications for the way data is stored and transactions processed,” said Schroeder.

Tech Disruptions Coming for FinServ, Insurance

Juniper also sees major technology-based disruption coming to financial services and insurance companies that have lagged behind the competition in adopting new technologies.

“Driven by the emergence of tech-challenger banks, the financial sector will witness increased disruption from technology-first players,” the company said. “Consequently, given the failure of many traditional banks to innovate and keep pace with pure play providers and technology banks, this will lead to a number of high profile acquisitions and consolidation within the sector, alongside continued direct investments. This will also be driven by the inability of challenger banks and payment banks to rapidly scale and maintain profitability compared to the incumbent players.”

Juniper also said the GAFA (Google, Apple, Facebook and Amazon) Group is investing in emerging fintech markets, such as payments (contactless and mobile). “It is our view that alongside banks, they could, and will, buy out start-ups to expand across other payment verticals, such as money transfer and remittances, to become more fintech-relevant.”

As for the insurance industry, Juniper said, new AI technologies are enabling insurance to become much more personalized. New entrants that have adopted “insurtech” have a more nuanced approach to insurance than traditional providers, with products that enable customers to tailor insurance to their specific requirements and allow more accurate calculation by insurers of risk exposure.

Juniper cited industry data from Accenture showing that insurtech investment rose from $800 million in 2014 to more than $2.6 billion in 2015.

“Traditional players now fear losing market share to new insurance tech start-ups, such as GetSafe,” Juniper reported. The company also “believes that if Google, Facebook and Amazon enter this market, leveraging their huge customer base, it would be a far more attractive value proposition for them than the incumbent players. While many of the world’s largest insurers (including Aviva, AXA, Allianz, AIG, MetLife and XL Catlin) have created venture capital funds to invest in promising technologies and propositions, Juniper expects more companies instead to partner with the insurtech firms.”

Pay Now, Pay Later

Several of our crystal ball gazers expect changes in technology purchasing strategies within the enterprise.

“IT spending has never been higher, but don’t expect to see as many long-term contracts with legacy players in this cash-infused, IT future,” said Ash Ashutoff, CEO of Actifio, a data virtualization company. “With startups continuing to accelerate innovation, customers will continue to look for new solutions more frequently, leading to shorter, fast ROI-driven IT purchases.”

Galton-Fenzi has a similar vision of IT consumption, contending that “businesses are less inclined to make the large, upfront investments we’ve seen in the past, something that’s reflected in the increasing popularity of cloud-based services and pay-as-you-grow subscription models.

He said this also applies to networking, particularly with the growth of software defined data centers.

“Customers are starting to opt for open stack, non-proprietary, software-based, cloud-managed solutions, and that involves a move from traditional perpetual licensing to an increasingly popular subscription-based, pay-as-you-go model.”

Avi Networks’ Pandey anticipates that apps in the era of digital transformation with impact traditional IT roles, including buying decisions.

“With companies rapidly managing their business and IT infrastructure with software-driven solutions and processes,” Pandey said, “application owners will have a greater say in IT purchasing and resource allocation to deal with issues including APIs, programmability and shadow IT. Similarly, infrastructure and operations teams will require greater visibility and troubleshooting tools to accelerate delivering apps from hybrid clouds and to optimize app performance.”

EnterpriseAI