By Sanjay Lodha, Chairman and Managing Director at Netweb Technologies
A recent KPMG global tech report 2023 surveyed more than 2100 technology executives from 16 countries and came up with some interesting findings. According to the survey, 29% of the businesses said they have observed a profitability or performance gain of at least 11% from the investment in data & analytics. As a part of the digital transformation strategy, companies continuously invest in technologies such as data & analytics, cybersecurity, public cloud and XaaS technologies, AI and automation, VR/AR (including the metaverse), and low code/no code.
Small and large data centers are the cornerstone for a successful digital transformation strategy, housing the infrastructure crucial for hosting business-critical workloads.
In India, with the rapid creation and consumption of data, the popularity of data centers has been on the rise. Furthermore, the propensity of citizens to access services via digital mode is accelerating the growth of data centers. In 2023, the total number of data centers in India is estimated to be 151. Before we discuss how new and advanced data centers – centered around GPU acceleration- are redefining the growth of our country, let us first understand the definition of data centers.
A data center is a network of storage and computing resources that facilitates the delivery of
shared data and applications. They act as the nerve center safeguarding massive volumes of
critical data and are crucial for the day-to-day functioning of companies and consumers alike.
Digital infrastructure stack and enterprise requirement adding impetus to the growth
As the country has developed a range of digital public infrastructure stacks such as Aadhar,
CoWIn, ONDC, and UPI, the ecosystem to store and save them is equally important. Identifying the future demands, the data center companies, both domestic as well as international, are coming together to prepare specific business use cases. Moreover, the growth at the enterprise level is fueled by organizations keen to adopt new game-changing technologies such as AI/ML, IoT, Cloud, Blockchain, Big Data, AR/VR, etc.
Modern enterprises pursuing rapid digital transformation to add value are digitizing their core services, leading to an exponential increase in data consumed and produced. In the coming decade and even now, a dominant portion of the services we access is expected to be digitized. Modern data centers incredible computational power is expected to improve the quality of service and our lives.
For instance, generative AI that relies on the cutting-edge hardware infrastructure of data
centers can generate 3D simulated environments. It can produce generated images of rural
areas, irrigation, and terrain. These depend on training the generative model for 3D environments such as Generative Adversarial Networks (GANs), and Variational Autoencoders (VAEs) on a large set of satellite images. It is significantly cheaper, more effective, and far more instantaneous compared to achieving the same by flying drones.
E-commerce. Cloud services and Data Centers
The success of the digital economy is positively correlated to the data center infrastructure.
Despite the global headwinds, the growth of data centers has been on an upward trajectory.
The expansion of e-commerce in India with stringent rules on the localization of data is
imparting momentum to developing data centers in India. The factors in which data centers
impact the agility and efficiency of e-commerce platforms are scalability, reliable uptime
(constant availability and disaster recovery), data storage and management, data security,
content delivery networks, and data analytics & AI support. The data center growth looks promising, with the e-commerce market expected to reach USD 350 billion by 2030.
Setting up a data center can be daunting for small and medium enterprises as it requires
enormous resources and technological know-how. However, such enterprises want to be
included in the technology race. Consequently, they are exploring different avenues of cloud
services from the data centers set up elsewhere. Enterprises can use different models
categorized by deployment (Public, Private, and Hybrid) or services (SaaS, PaaS, or IaaS) based on requirements. Cloud services fundamentally rely on data centers, as the latter forms the physical layer where the virtualized services operate. Cloud service adopters are looking to leverage the best IT resources and desire dependency, diversity, and efficiency.
Furthermore, the wonders of edge computing have accelerated the adoption of cloud services as enterprises have a strong opinion on the collection and processing of data near to the customer. The Indian cloud computing market is projected to grow at a CAGR of 18.37% between 2023 and 2029. The market size was estimated at USD 6.47 billion in 2022.
Digital India, Make in India: The promising road ahead
Thankfully, the GoI has been proactive in recognizing the role of data centers in driving the
agenda of public welfare schemes such as “Digital India” and “Make-in-India”. Like private
organizations, the government also generates a humongous amount of data that needs to be
managed, processed, and secured, as it contains confidential information about citizens. To
achieve these objectives, the government must ensure that data centers work optimally. The
government’s bullish approach to promoting high-performing data centers is expected to
attract more investments, shaping a data-driven economy.
The Government of India launched the & Make in India& initiative to encourage companies to manufacture their products in India and incentivize dedicated investments in the manufacturing sector. This initiative has positively impacted the growth of data centers in India for several reasons, such as policy support (Incentives, Infrastructure Development), increased local demand (Digital transformation, E-commerce boom), data localization laws, investment opportunities, and technology adoption.
With the rollout of 5G services, the accuracy and speed of delivery of public services is expected to increase significantly. As consumers will look to avail more services, there will be the need to store and process such data in advanced data centers.
To democratize the benefits of new and emerging technologies, including generative AI, it is
imperative to have modern data centers with advanced software and hardware capabilities.
The debate around building a sovereign GPU cloud is gaining momentum. Generative AI models are advanced neural networks that need enormous parallel processing capabilities from the processors that run them. Graphical Processing Units (GPUs) play a crucial role in the optimal training of AI learning model that uses parallel processing to yield the desired result.
However, the traditional data centers that have been working with CPUs will still be relevant and used for inferencing purposes with structured data. The AI learning market is already gaining traction and shows no signs of slowing down. With the steady growth of the AI learning market, the inferencing market is also expected to increase in size as the actual delivery of the solutions starts accelerating the demand in the existing supply chain. Therefore, even with the GPU acceleration, the requirement for CPU will still be very pertinent.
The switch to a GPU data center is expensive. However, if we evaluate the cost/benefits
analysis, it is worth the gamble. The popularity of GPUs has evoked unprecedented interest
from both established players and new ones. Although at this stage, the incumbents have a
slight edge, it would be interesting to see how the market turns out in the coming years with
the rapid adoption of GPUs in modern data centers.
Conclusion
Summing up, the future of the growth of data centers looks perfectly on track with a
continuously evolving ecosystem. India is uniquely poised to make the most of this digital
transformation opportunity, and data centers built around advanced GPU acceleration
capabilities will change how we are perceived as a technology laggard.
