Thriving In The Data-driven, AI-led World

The recently held Data Driven World event by Raritan in association with BusinessWorld focused on the contours of a data-driven and AI-led world and the IT infrastructure choices available to lead and thrive in a continually disruptive environment. Anjani Kommisetti, Country Manager-India and SAARC, Raritan and Servertech, explained that the event series that has been running for many years is designed to get the thought process and feedback from the industry about where IT infrastructure is heading. 

The infrastructure decisions to lead and thrive in the data-driven and AI-led world are but one way of future-proofing the organization for the evolving digital world. 

Driven by Data, Led by AI

Outlining the evaluation of data-driven business models in various industries, Vidyasagar Zakkula, Software Integration Head, AMD, illustrated the case study of how IoT devices and hosted cloud solutions can be used in the healthcare industry. How the IoT data streams can be aggregated at in-premises non-conventional micro datacenters with edge computing and regular datacenters to formulate the global strategy, common platforms for various players in the ecosystem. How various AI models and analytics can run to share the information with various third parties for the best of the healthcare industry, and more. 

Zakkula’s point is that for such a smart healthcare system to work seamlessly, there has to be a structured approach to managing data and IT infrastructure components that align with this structure. 

Next is AI. The march of AI has been relentless, and the trend is irreversible, notwithstanding the barrage of criticism on the dangers of AI. Microsoft has been one of the forerunners of AI both in consumer as well as the enterprise space. It has also been leading the use of AI in social transformation. 

Dr. Rohini Srivathsa, National Technology Officer and Strategy Lead, Microsoft India, outlined the company’s vision for AI – putting it to work for the good of business and humanity. Microsoft’s approach to AI comprises three pillars: a) leading and enabling breakthrough innovation that can be scaled at speed; b) putting people at the center -  empower people, every part of the ecosystem to use AI for positive change and c) using AI responsibly for the greater good of humanity

She underlined the fact that driving innovation is core to Microsoft AI. The company has been investing in the key building blocks and the last couple of years alone have seen multiple breakthroughs with AI achieving human parity in vision, speech, machine reading and translation.

The next logical step is to place these breakthroughs in the hands of developers to build on. You must provide a powerful platform that accelerates innovation by developers and partners and makes it more accessible. This in turn helps businesses transform by applying the power of intelligent technology both to current approaches and entirely new scenarios. Dr. Srivathsa pointed out that 1.3M developers are currently using Azure Cognitive Services. She also underlined the need for businesses to identify and bridge the gap between the desire to use AI and the skills and strategy needed for successful deployment.

But it isn’t about business alone. The promise that AI holds out is what it can do for each of us. Rohini emphasized the need to democratize AI and make it accessible to all. The full potential of AI will be realized when it is taken to the grassroots for inclusive growth and societal good.  It is already helping transform healthcare from being preventive to predictive. AI-powered smart agricultural practices are increasing crop yields in the country.

Dr. Srivathsa concluded by reiterating that ethical designing and deployment of the technology is a must and that a dynamic infrastructure combined with a collaborative ecosystem is paramount for AI to work for India.

Making the Right Infrastructure Choices

The event featured a panel discussion on the topic: Building Datacenter for Unpredictable and Dynamic Future. The panelists were: Satish Vishwanathan, Chief Product Officer, NxtGen Datacenter and Cloud Technologies; Raghavendra Vaidya, Sr. VP-IT, Mercedes Benz India;
Sudhakar Gooty, Global Head- IT and Datacenter Products, NetRack and Ranendra Dutta, CTO, Terrago Logistics.

Dayaprakash, Contributing Technology Editor, BusinessWorld who moderated the panel discussion set the context by identifying three key driving trends: 

  1. Data is the new oil— We now realize the value of data and we are looking at harvesting data in all possible avenues. Data comes in from personal devices, machines (IoT), satellites, surveillance, software, networks, applications, cloud ecosystems, and computer systems. Analytics, machine learning, and AI are applied to data. 

  2. Customer is the king— customer experience is the topmost performance indicator, which in turn is determined by speed of response, accuracy of response, quality of response, and adaptability to change. Customers could be internal and external. In a business, the business teams are customers for the IT organization.

  3. Disruption would continue. Every wave of technology presents a disruption for business which translates into a disruption for IT. IT should be able to handle these disruptions in increasingly shorter time periods, at costs that business can handle, and ensure safe and trustful computing. 

Each of the above three areas has enormous implications on the IT infrastructure, which is largely typified by the datacenter and cloud. The topic of the panel discussion centered on making the right infrastructure choices against the backdrop of the three trends above. 

The discussion elicited a broad range of views from the panel comprising veterans in the datacenter industry and IT leaders in the enterprise. 

Datacenter Evolution- Seeking Efficiencies and Being Adaptable

Tracing the evolution of datacenters, Gooty said that while every tech disruption had an implication on the IT infrastructure set-up, CIOs are looking deriving more business avenues and efficiencies from technologies that they are already invested in. That is, the transition is never a ‘cutover from the old- to- move to the new’.

Satish Vishwanathan traced the evolution in terms of better efficiencies in terms of operating costs. He alluded to the phase of datacenter tiering dictated by differing demands in terms of adaptability and efficiency. He said that datacenters are the biggest consumers of power— a good-size datacenter consumes 100 MW of power and power accounts for nearly two-thirds of the total cost of operation. 

Satish brought up another interesting aspect about the change in business model that necessitated the active use of IT infra from working business hours earlier to 24x7 in the digital age. 

Complexities in Innovation

Raghavendra Vaidya commented that the future cars are like ‘datacenter on wheels’ which has around 95 computers inside it. Sharing from his past experience in various global organizations, he said that innovations were driven not by making the hardware better but working on the software instead. He also added that innovation cycles were getting shorter and more complex, which could be a problem from a business point of view as there is a cost incurred in working on these innovations in infrastructure.

Pace of Disruption is Slowing Down IT

Ranendra chimed in with the point that with the increasing pace of disruption, IT is not able to perform at the pace business demands due to change in business models. Business would hear about the latest new technologies in digital transformation activity and ask IT to deliver it tomorrow, but IT has numerous considerations to take care of and cannot jump into the new thing. The other difficulty is that business requirements don’t come in a structured manner; IT is left to figure out the pieces.

Agility with Cloud and Computing-at-the-edge

Satish is convinced that cloud has allowed people to look at different business models. He gave the example of a mobile manufacturer who could liquidate the entire inventory in 15 minutes through online sales only because of the elastic infrastructure that was available on cloud. He predicted that edge is going to become mainstream. 

There are many datacenter formats now available including datacenter-in-a-box, most running on commodity hardware, and even self-cooled datacenters. IDC estimates that by 2021 the world would need nearly 20,000 edge datacenters. 

Experiment Continuously and Keep Choosing the Best Models

Raghavendra Vaidya spoke about how infrastructure choices that were made in the past would have a bearing on the future. In the pre-cloud era, well-established companies took pride in engineering their own datacenters. This continued until the initial years of cloud. But soon, it was becoming difficult to keep up with the innovation on public cloud and therefore the decision to shift.

Despite the marketing hype that cloud is cheap, the reality is that unless the applications are built for the cloud, running legacy applications on public cloud would be phenomenally expensive. Thus, they decided to use a hybrid approach of using a mix legacy, private cloud and public cloud to run the applications best suited for the respective models, adding newer points as they emerge and continuously evaluating and shifting applications to the best suitable model.

Raritan Viewpoint

IT Infrastructure for Digital Transformation
Most new datacenters operate at optimal availability and with infrastructural energy efficiency close to theoretical design targets. As such, it might be argued that the two biggest challenges of datacenter technology in the past 30 years have been addressed. But despite this progress, the pace of change in the datacenter industry will continue and is likely to accelerate over the next decade and beyond. This will be spurred by increasing demand for digital services, as well as the need to embrace new technologies and innovation while mitigating future disruption. At the same time, there will also be a requirement to meet increasingly stringent business parameters and service levels.

Speaking at the forum, Anjani said that datacenters have continued to evolve over the years, but it is still difficult to predict the shape of things to come. He alluded to a few key trends that IT leaders need to consider while making infrastructure choices:

1. Hardware is highly configurable unlike in the past. This opens out more choices and easier options. 

2. Small datacenter boxes are now available. They could find applicability in many areas. 

3. Racks are becoming denser in terms of power and compute capability. A single rack solution can handle from 3 KW to 55 KW. It offers a huge swathe of application areas. 

4. Cooling will go away from datacenter at least from edge datacenter.

5. Whiteboxes are a great option for all datacenter needs because it allows building solutions for dynamic requirements. 


dummy-image

Ed Nair

BW Reporters The author is Consulting Technology Editor

Also Read

Stay in the know with our newsletter