Given its population of more than 1.3 billion, India is a vast market for any industry. The data centres (DCs) sector is no exception. As per an Anarock-Mace report, within the next three years India will witness the construction of 28 hyperscale DCs at least, across more than 16 million sq. ft., having an IT power capacity above 1,400 MW.
Since the Centre is keen on promoting its Digital India mission, the government seeks to attract more investments in DCs to realise its emergence as a data-driven economy.
A JLL report notes that India’s DC capacity is slated to rise from 375 MW in 2020 to 1,078 MW by 2025, representing a $4.9 billion investment opportunity during this period. The sustained demand from investors is expected to be propelled by the latest data protection laws, edge computing, the IoT and a shift from captive to colocation DCs as well as 5G, during these five years.
Not surprisingly, the world’s second-largest DC was inaugurated in Mumbai in July 2020. Even before the coronavirus outbreak, the demand for cloud solutions and DCs was already rising due to data localisation and other imperatives.
The pandemic has only accelerated this demand, intensifying the need for hyperscale and edge DCs in the new normal because more businesses have begun operating remotely. A Crisil report reveals that data consumption in FY2021 shot up by 38% year-on-year because of the pandemic.
With businesses using greater amounts of data every year, managing this is crucial for growth. DCs cater to this demand by offering data backup and recovery facilities, simultaneously supporting transactions and cloud storage applications.
According to market intelligence company International Data Corporation, a DC turns ‘hyperscale’ if it crosses 5,000 servers and 10,000 sq. ft. In the past two years, India has seen soaring demand for hyperscale DCs as companies have been moving IT infrastructure to the cloud, nudged by multiple drivers, including the mounting requirements of OTT platforms and app-based services.
The impending 5G rollout is anticipated to push greater demand for DCs as it would introduce new applications calling for high reliability and low latency.
With most businesses migrating critical data to the cloud, latency becomes an issue as a faster response is inhibited when the workforce is distributed over a large area. But edge DCs can boost network performance by largely reducing latency via real-time data processing and swifter responses. By processing data at the edge,unlike centralised or distant DCs, edge DCs provide more efficiency.
Meanwhile, in improving efficiencies and boosting productivity, artificial intelligence is now playing an increasing role. As the world witnesses an exponential surge in data annually, it is placing immense pressure on DC infrastructure, calling for more complexity.
But as DCs turn more complex, it is difficult for human resources to deal with the greater complexity without affecting efficiency and performance levels. This is where AI-enabled DCs can help organisations in significantly enhancing efficiencies.
Google highlighted the benefits of using AI in improving the energy efficiency of its DC. Within barely 18 months, Google’s AI-powered system ensured a 40% reduction in cooling energy consumption. This represents a 15% drop in its overall PUE (power usage effectiveness) overheads.
By augmenting the intelligence and automation of equipment and management systems, DCs can be more efficient and reliable in both energy usage and operations. By optimising power consumption, electricity costs can be reduced, which is a vital factor in DC infrastructure.
In recent years, major developments in AI and machine learning (ML) have helped in augmenting DC facilities. Therefore, DC algorithms built for task automation and predictive maintenance have become more refined, permitting IT heads to spend less time on routine tasks and more on future planning.
This is possible as these algorithms use empirical evidence or data in predicting more accurately when maintenance may be required. In this way, data-driven predictive maintenance models can alert IT personnel if something is about to fail, limiting the possibility of failure.
As an advanced subsection of AI, ML scrutinises and pinpoints patterns in huge stacks of data. In this manner, it can optimise all aspects of DC operations such as designing and planning, managing IT workloads, uptime maintenance and cost control. The IDC report states that half of the IT assets in DCs can run autonomously via embedded AI functions.
Moreover, AI-enabled DCs can tremendously boost data security since such centres may be vulnerable to cyberattacks. AI achieves this by detecting abnormal behaviour in networks or deviations in the regular pattern to thwart cyber threats.
The coming years will see a significant jump in the number of IoT-enabled devices, thanks to 5G, which canmajorly minimise latency. Additionally, India’s mission of emerging as a digital economy will generate massive mounds of data in e-commerce, online education, digital healthcare, live–streaming entertainment and more.
All of which will require resilient, agile and scalable IT infrastructure. In such a scenario, data centres will play a pivotal role in powering India’s Internet-enabled, data-driven economy.
- Venkatraman Swaminathan, VP & Country General Manager, India & SAARC, Secure Power Division, Schneider Electric