As India moves to being a digital economy, in the last couple of months, three conglomerates, Reliance Jio, Adani Group, and Hiranandani Group, have announced plans to set up data centers in India. While Jio, over the next 10 years, plans to build data centers across India that run Microsoft’s Azure cloud, Adani expects to invest `70,000 crore to build data parks in a southern state over the next two decades. And the Hiranandani Group is looking to invest about `15,000 crore over the next 5–7 years in setting up data centers in Mumbai, Panvel, and Chennai.
“The government is looking at data localization. If that happens, it will lead to exploding data center requirements. This will bring in the Googles and Amazons of the world.”
International players like Bridge Data Centers, Colt Data, Equinix, Cyrus One, and ByteDance, the owner of TikTok, have also started investing in India. Players such as L&T, Shapoorji Pallonji, Epsilon, and Pi Data centers and organizations in the traditional construction business will also make inroads in this space. Indirect investment will be huge in terms of technology and automation as well. It is expected that various data center products will also see an increase, with servers expected to see maximum traction.
There is a government push in India for digitizing of business and services, with multiple states also looking to set up data centers. There is also a need for greater data localization to protect citizen’s digital assets. The recently introduced stringent government regulation on data privacy that requires foreign companies to store data within the country is also providing impetus to this sector.
According to ResearchandMarkets, the Indian data center market, on an annual growth rate of almost 26 percent is expected to reach `28,000 crore by 2024. There are 540 million active internet users in India, but the country has an abnormally low data center capacity of less than 700 MW, compared to Europe’s 8600 MW capacity for 460 million Internet users, says data center map. Bank of America Merrill Lynch estimates that in the five metros itself, India, with an investment of `52,000 crore, will build an incremental data center capacity of 700–800 MW IT power in the next five years.
The Asia-Pacific includes a more diverse range of data center markets than any other region of the world; therefore, predictions need to be made with some caution. DatacenterDynamics has done a detailed study on this region.
The gradual evolution to digital transformation within data centers is taking shape fastest in the markets of North-East Asia, China, and India. Singapore is also joining the process. India is joining China in developing a critical mass of technological innovation, supported by the necessary capacity of digital infrastructure to move into a regional leadership role.
At the moment, the early adopters are taking shape and the move toward AI and machine learning (and toward Edge also) occurs in a small range of companies in what appears to be a developmental capacity. The future will continue to arrive through 2019 but not in full force.
In China and India, the future will take in the two extremes – hyperscale on one hand and Edge in the longer term on the other (and the two in co-existence). The expansion among the upper end of cloud and facility service providers will continue to put resource pressure on the hub markets of Singapore and Hong Kong. Expect growth to continue in areas immediately around Singapore as well as in markets of North and Eastern Asia adjacent to China.
Technology is also being asked new questions more directly. The reasons for deployment and change are more emphatically answerable to business than they have been before, and the direction of future technologies appears to be as much about business as technology. In a sense, it always has been – operational costs emerge as a key concern as they have done in every DCD survey for the past 15 years. But now, the speed of technological evolution and with which business and individual consumers catch up and create demand, has forced the issue into the open. The issues of connectivity and virtuality (=the non-physical), which will continue to challenge security (a continuing major concern) and legislation, are challenging also the concepts of investment and return. The road to the land of opportunity via IT is open but it is not always clear who will maintain the highway or pay the toll.
Strong increases in both private and public cloud everywhere across the region and increasingly in hybrid IT as well means these issues of investment, security, and compliance will add to the pressures on those owning, managing, and legislating for data centers and networks.
This means that the roll out of digital transformation and the impacts on infrastructure will increasingly take the form of partnerships not just between innovators, suppliers, and project visionaries but between different regions and types of community – urban > rural and remote in China, India, and Australia.
Challenges to the region’s data centers
Rising operational costs are the key challenge to these APAC data centers. This is true equally across the two main categories of data center (enterprise and service) and it is nominated 75 percent more than the second-most nominated reason, which is the challenge of capacity planning. Concerns about threats to cyber security, costs of facility upgrade, dealing with an onslaught of new technologies, and about risk management round out the challenges most nominated.
Among end-users, data center concerns are higher about cyber security, risk management, and the cost of upgrade. These indicate the awareness of impending transformation and pressures on demand this will bring to many of these data centers. Service data centers (colocation, cloud, hosting, and the like) unsurprisingly face greater challenges in meeting the needs of diverse clients and also from rising power densities associated with the transition toward higher demand services.
There is some variation by market
Australia and New Zealand are generally close to the average but face particular challenges of compliance (28 percent).
China is below average on most concerns except risk management (34 percent).
Hong Kong follows this pattern with very much lower concerns about cyber threats (11 percent) and capacity planning (21 percent)
India presents the picture of a market preparing for increased demand and technological transition with higher nominations for the challenges of new technologies (33 percent), the cost of upgrade (30 percent), the separation of FM and IT in data centers (21 percent), the problems of integrating equipment from a variety of vendors (20 percent), the hosting of hybrid IT (28 percent), and the evolution of Edge (14 percent).
Indonesia also profiles as a market undergoing structural change with particular challenges of capacity planning (50 percent), hybrid IT (30 percent) and digital transformation (34 percent) – all three the highest in APAC.
Singapore is challenged considerably more than any other market by operating costs (70 percent), and by high levels of concern about cyber threats (44 percent), compliance (31 percent), and the demands of digital transformation (25 percent).
The markets of South-East Asia and Indochina indicate concern about risk management (38 percent), skills shortage(25 percent), and IT/ FM separation (25 percent); those of North-East Asia indicate higher concern about the costs of data center upgrade (44 percent) but lower than average concern about operating costs (36 percent), cyber security (20 percent), and hybrid IT (4 percent).
These challenges have some capacity to alter the Asia-Pacific data infrastructure ecosystem; as the in-house data center is no longer the default option, other means of data storage, access, and processing have arrived.
From data center to digital infrastructure
The shift away from in-house data centers since the research of 2017-18 has been fairly emphatic – from 53 percent to 49 percent. While the shift is notable, some of it will possibly be accounted for by the level of variation between two different sample measurements as well as increased sample representation in 2018-19 from markets with lower levels of in-house data center deployment – Singapore and China in particular. Proportions in colocation have increased over the past year as has the proportion in both private and public cloud.
The profile of footprint indicates a divergence between market segments. China and Singapore are the first group. They show lower dependence on in-house and the highest levels of dependence on both colocation and cloud. This situation is made more possible through the influence of centralized planning on data center development and in both cases it is seen as necessary in the conservation of energy, water, skills, and money and, in the case of Singapore, of space as well. The second group comprises the older established markets such as Australia, New Zealand, and the markets of North-East Asia, which have a more average balance between the options, and where the in-house option is used at the center of an outsourced and cloud portfolio.
The third group includes the more recently evolved markets of India, Indonesia, and South-East Asia. These have the highest dependence on in-house, and their uptake of cloud, services, and outsourcing options is more patchy and more dependent on what options are available and trusted.
Over the next few years, this sample will see moderate reductions in the use of in-house data centers and also in the use of other outsourcing options (managed services, hosting, and more). This latter trend may be temporary rather than long term, but it would suggest a level of facility services that are being bypassed by cloud as part of the broader evolution out of physical data infrastructure. It may also be due to the evolution of colocation providers toward IT and cloud services, away from their heritage of facility services.
There will be a 6 percent sample increase in the use of colocation facilities and services and higher (weighted by footprint) increases in the deployment of private (20.7 percent) and public (14.1 percent) cloud.
The strongest levels of increase in these digital infrastructure options occur in the major growth markets of China, India, and Indonesia. Growth rates of these markets are higher than 20 percent for private and public cloud as well as double-digit growth for colocation and even very limited growth in in-house to develop hybrid or as companies get large enough to justify investment in their own facilities.
The key reasons for migrating between data infrastructure options include the need to ensure data security (both physical and cyber), guaranteed availability, cost, and scalability. Interconnectivity, service levels, availability, and access to networks/carriers are key reasons for moving from in-house into colocation; increasing agility, scalability and pay-as-you-go for migrating to cloud, especially public cloud. Investment in-house (and to a degree in private cloud as well) has a greater foundation in ease of management, compliance, and security.
The technologies and processes classified under digital transformation require data centers and cloud in order to deal with the data generated and the intelligence made possible. Yet transformation is pervasive, and it has the power to disrupt data center operation.
Transforming the data center
The introduction of digitized data gathering, analysis, and application technologies occurs within a narrow band of companies, which account for 15 percent to 25 percent of the overall sample. Fairly much the same set of companies are deploying across these four technologies and will continue to do so.
The profile between different data center types is fairly even, although big data/analytics deployment is higher among end users (29 percent), and data centers run by IT and managed service companies are slightly further advanced in terms of AI and machine learning (25 percent) and software-defined infrastructure (27 percent).
Challenges of Data Center Operation (% Nomination)
Companies with larger data infrastructure portfolios are more likely to continue to invest in future in three of these four technologies:
- Software-defined infrastructure (average 950 m2 footprint versus 347 not)
- AI/machine learning (824 m2 versus 384)
- Big data/analytics (836 m2 versus 377)
The geographic profile of companies which have, or which will deploy these technologies tends toward the markets of North-East Asia and China, and for software-defined infrastructure toward Australia and Singapore as well. Future uptake switches
One in four of the sample indicates a strong interest in Edge; a further 31 percent indicate some interest, and 12 percent little or no interest. The remainder are unsure. Interest is highest in China (31 percent) and India (29 percent).
Those who are most interested in Edge see it as the key to digital transformation, as a means of reducing latency and delivering better to customers. Cost issues do not distinguish between different levels of interest, which suggests it is the perception of ROI which is more important.
Those who are unsure or are negative about Edge are most concerned about the lack of definition. It is notable that those most positive to what they know of Edge have high levels of concern about it. This suggests that the levels of interest are currently formed at a level of business engagement and that those most interested may be happy to fill in the detail as it emerges. Yet without that engagement there is simply indifference.