Photo: Bloomberg
As Covid-19 increasingly forced workers out of the office and back to their homes, cloud computing has experienced exponential growth to ensure business continuity under lockdown. Amazon Web Service’s revenue 3QFY2020 ended December 2020 grew by 29% y-o-y to US$11.6 billion ($15.4 billion). Microsoft Azure reported 48% y-o-y growth in 1QFY2021 ended June.
This trend of growing cloud usage shows no signs of slowing down. An International Data Corporation (IDC) report found that 90% of global enterprises will rely on multi-cloud by 2022. In China, the Ministry of Industry and Information Technology forecasts its cloud market to enjoy a 30% CAGR from 2019 to 2023.
But what happens when existing cloud networks begin to run out of space? “As millions of workers shifted out of enterprise hubs into more remote locations, they put a strain on networks,” say Dave Russell, vice-president of Enterprise Strategy, and Beni Sia, vice-president, Southeast Asia and Korea, at Veeam. This creates increased latency and a need for more computer power, capacity, and storage.
The solution is to invest in edge computing solutions to reduce the strain on these overburdened networks. This involves bringing computational, data storage and connectivity resources closer to the locations where they are required to save bandwidth and reduce response times. Analysts expect edge computing to rise 30% a year to US$44 billion by 2030.
Teresa Tung, chief technologist at Accenture Cloud First, notes that edge computing is always on and enjoys low latency due to less network uptime, round-trip times and bandwidth constraints. Firms also save money as processing and storing at the edge of a network tends to be cheaper than doing so at its centre. McKinsey has identified 107 use cases for edge computing from energy management to real-time tracking of work-site safety.
One less well-known use case of edge computing is perhaps its ability to deliver services “quickly with a personal touch”. Restaurants, for instance, can adjust and optimise the performance of smart kitchen equipment based on real-time customer traffic data. Russell and Sia also highlight how edge computing can be used to provide customised retail experiences to customers such as through personalised coupons, promotions and shopping assistance.
“You can almost equate business resiliency to how an organisation handles, manages and leverages data,” Sia tells The Edge Singapore. In the face of a massive burst of data, he says, businesses must make use of technology to better manage and leverage this sheer mass of information to their advantage. Investing in edge allows for companies to maintain business continuity and the seamless transfer of data even in times of disruption like Covid-19.
Democratising computing
So how does edge computing work? At the heart of an edge network is a centralised cloud server, which has the highest density of computing, storage and networking resources. Besides taking on the most intense big data processing, it also performs data warehousing functions to ensure that the data is stored efficiently. It is the cloud that keeps the entire network running.
Under a traditional network architecture, all computation is conducted by the centralised cloud. An edge computing architecture, however, uses smaller distributed data centres known as edge nodes. These perform lower latency computing that provides real-time data processing, data visualisation and standardisation and control responses.
Relying on edge nodes brings computation and data processing closer to where data is being gathered — the edge devices. These include smartphones, laptops or internet of things (IoT) sensors. It is the greater proximity to these devices that brings about reduced latency. Trying to have multiple devices — each generating considerable amounts of data — transmitting to the cloud directly puts significant strains on bandwidth.
“An edge gateway, for example, can process data from an edge device, and then send only the relevant data back through the cloud, reducing bandwidth needs. Or it can send data back to the edge device in the case of real-time application needs,” explains journalist Keith Shaw in Network World.
This more cost-effective architecture can contribute towards the democratisation of technology and, by extension, associated commercial opportunities. Edge computing’s ability to aggregate, process and analyse data locally can help bring new business opportunities outside major commercial hubs. Savings in terms of data management cost, startup and operational costs as well as a larger talent pool can also reduce barriers to entry for new ventures.
See also: Quantum computing to edge near mainstream this year: Dell
IT research firm Gartner has identified the democratisation of data as one of the top ten strategic trends for enterprise in 2020. The shift to remote working has only accelerated this democratisation. “Home-based businesses will create their edge outlets, and organisations will establish regional edge hubs to serve growing numbers of remote employees and contractors,” predict Russell and Sia.
Securing the edge
Edge computing also provides cybersecurity benefits. Edge computing’s decentralised nature means that if a device is breached, the compromised endpoint can be isolated to prevent further spread — think about amputating a poisoned limb to save the rest of the body. More data can be kept at endpoints to limit the amount of information sent back to the centralised cloud, adding an extra layer of security near the network’s centre, where more mission-critical data is stored.
But edge networks can also mean a larger attack surface for bad actors to exploit. “Adding more nodes opens up more places vulnerable to attack. This will force IT security leaders to bolster their defences to ensure that information and applications stored at the edge match the strength they are applying inside the data centre itself,” warn Russell and Sia.
The relative novelty of edge architecture, Sia observes, can put cybersecurity experts “in a little bit of a foreign land”. A survey by tech firm Kollective found that around two-thirds of IT teams view edge computing as more of a threat than an opportunity.
To combat this, Russell and Sia advise firms deploying edge computing to manage edge devices properly. This involves establishing policy-driven configuration enforcement and security for computing and storage, with encryption of data essential whether it be static or being transferred. Most edge providers, says Sia, do not provide cybersecurity or even operationalisation solutions with the platform, leaving users to secure the networks themselves.
“Businesses must be mindful about which suppliers they choose — consumer-grade tech is often a no-go for a start, which is a challenge for smaller firms where one in four use it for their data security. They also need to stop putting total trust in basic endpoint security measures like passwords and perimeter defences,” notes a thought piece by cybersecurity firm Kaspersky.
But the weakest link in cybersecurity is always the human users. Firms, says Ken Mafli, senior digital marketing manager at Townsend Security, should therefore adopt a “zero-trust” approach to computing where each user should be restricted to the least amount of data needed to do their jobs. Users should be challenged each time they log in to prove they are authorised to access the network and that their network is secure.
Yet, as important as cybersecurity is, says Sia, it is only but one of the problems encountered when implementing edge computing. Ultimately, he says, security cannot come at the expense of efficiency and cost-effectiveness — key attributes that enable an edge to deliver the desired business outcome. Businesses in Asia are already struggling to integrate edge computing into legacy systems as they transition towards this new type of network architecture.
Edge in Southeast Asia
The Asia Pacific is an especially lucrative market for edge computing, seeing that it is the leading region for digital transformation in front of Europe and the Americas. Future Market Insights finds that Asia Pacific is the largest revenue contributor to the global smart factory market — a key user of edge technologies — accounting for US$20.4 billion in 2014. Spending on IoT in the Asia Pacific, IDC predicts, could reach US$288.6 billion in 2021.
Southeast Asia is an interesting market due to its interest in developing “smart cities”. The Asean Smart Cities Network sees member states seeking to use “smart technologies” to promote sustainable and livable cities as the region’s strong demographic and economic growth speeds up the pace of urbanisation. Covid-19 saw the region welcome at least 40 million new internet users, with 70% of the regional population — or approximately 400 million people — now online.
“The Edge Cloud and IoT revolution may be taking the global IT industry by storm, but it’s South East Asia that’s expected to see maximum action,” notes a thought piece by Kshitish Purohit, chief product officer at Indian cloud solutions firm Indiqus.
But Southeast Asia faces infrastructure gaps, says Justin Chiah, senior director for Southeast Asia, Taiwan & Hong Kong at Aruba. This is especially due to the difficulty in deploying broadband infrastructure on island states like Indonesia and the Philippines, which increases the cost of establishing such infrastructure. The Asian Development Bank sees the region needing infrastructure investment of US$210 billion per year to maintain growth momentum.
Building an edge hub
Now at first glance, edge computing seems to be more relevant for larger countries like Indonesia and China rather than a small island like Singapore. But Marc Wee, country manager for Singapore at Aruba, says that there remain significant opportunities for local enterprises to unlock its value as a business asset.
“With edge technology, organisations can process, store, analyse, and act on the data to optimise their business model and develop innovative products, services and experiences that will transform their offerings for customers as well as employees,” Wee tells The Edge Singapore. With edge computing key to digital transformation, this technology will be crucial to help Singapore realise its goal of becoming a “Smart Nation”.
In terms of data centre set-up in Singapore, says Sia, the government has put a lot of investment into this space to develop system technology. There are currently many systems available to businesses to ensure that their systems are as efficient, agile and light as they need to be. According to the Arcadis Data Centre Location Index 2021, which surveyed 50 cities, Singapore is the second most attractive city to build data centres worldwide.
Steps have been taken to streamline cloud computing processes with a strong regulatory framework. Infocomm Media Development Authority (IMDA) launched the Singapore Standard (MTCS) in October 2013 - the world’s first cloud security standard that covers multiple tiers of cloud security. Singapore also issued the Cloud Outage Incident Response (COIR) guidelines in March 2016 to help businesses mitigate the fallout of cloud outages.
Money, Sia adds, has also been poured in to develop technology talent to best take advantage of such emerging technologies. In 2019, IMDA and Microsoft signed a Memorandum of Intent to accelerate the development of Singapore’s 5G ecosystem. Edge computing is one proposed area for collaboration.
“IMDA looks forward to more partnerships with enterprises, technology companies, research institutions and Institutes of Higher Learning in our efforts to drive 5G innovation to support Singapore’s digital economy,” said Tan Kiat How, chief executive of IMDA.
“What Singapore needs to do well is to have that credibility,” Sia tells The Edge Singapore. Highlighting the need to build trust in edge technology, the integrity and availability of data in Singapore’s cloud networks must be maintained to ensure that trust remains. With its long track record of good governance and a strong commitment to digital transformation, Singapore is headed in the right direction when it comes to developing edge computing.
