“Small Data Centers Will No Longer Exist”

Unless they have a niche value proposition, says Bishwanath Ghosh, CIO- Corporate, Mahindra & Mahindra. Very large data center companies will rule the game while user IT companies will focus on app-level clouds

“When I started my career, there was nothing like a data center. You had a computer room which housed a large mainframe. The room would be very chilly, almost freezing! You could look at the mainframe from a distance, but you didn’t really know what it did,” says Bishwanath Ghosh, CIO- Corporate, Mahindra & Mahindra (M&M).

“You would leave your stack of punch cards with the data processing professional, who would return the output cards after two days. And it mostly showed the syntax errors you had made! I am from a Mechanical Engineering background, so my work would get lower priority than the computer engineers,” he says with a smile in his voice.

It’s a story that many seasoned IT professionals would immediately relate to.

After having spent close to three decades setting up and running the IT for automobile manufacturers to deliver business value, Bishwanath leads M&M’s IT initiatives today. In 2013, he received the ‘Best IT Driven Innovation in Automotive Business’ award from NASSCOM for eight innovative projects for auto and farm equipment sectors of Mahindra.

From running Fortran and Pascal programs three decades ago, to leading IT projects that plug into business models, Bishwanath has been working at creating business applications to enable strategic differentiation for the organization.

It’s a Sea Change

Bishwanath reminisces of the days gone by. In mid-1980s, more mainframes started coming in, he says. Then came the server room with terminals around in close vicinity. More storage came in, and backups started happening. Then the data centers came into being. Back then there were no network rooms, as connectivity was extremely limited. Security was not an issue because there was no connectivity. You had 64 kbps lines, for which you wouldn’t even know the real throughput!

When client-server architecture came into play, the data centers that we see today started shaping up, he continues. The computer center morphed into the data processing center, then the electronic data processing center. Information Services morphed into the IT departments we know today.

PCs became mainstream, IT usage increased, more people got connected. By mid-90s, email had come to India. Connectivity was very expensive. “In 1996-97, a 64 kbps line was a prized connection. A 14.4 kbps Internet connection used to Rs 25,000 back then!” he says.

The entire scenario has changed now, says Bishwanath. The ability to connect has changed the way data centers are perceived today. From single servers, people moved to multiple servers and that changed everything. “The conversations on how much to invest, how to build redundancy and how to have disaster recovery started happening at the turn of the century,” he says.

Businesses today are completely dependent on IT systems. “A couple of decades ago, work would not come to a standstill just because IT systems were not available. Today many companies have to stop their manufacturing operations if the applications or systems are down,” Bishwanath says. Today you need reliability—of data centers, applications, networks, end devices, power, air conditioning, basically everything. “Life is far more difficult now,” he says with a smile.

The Challenge: Keeping Older Technology Relevant

What are the difficulties he faces, we ask.

“We have our own data center that runs the key, critical systems for us. Any element that we purchase needs to have a life of 5-6 years. But technology changes faster, you get better systems at lesser prices. The key challenge for us is to keep every element relevant and efficient,” says Bishwanath.

“Another challenge for companies that run their own data centers is getting multiple power supplies from different energy providers. Terminating network also remains an issue,” he adds.

What about people challenges, we ask. “It’s not a challenge for user IT organizations like ours,” says Bishwanath. There are many good service providers who can run the data center well, he says.

Data Center Consolidation Up Ahead

We ask Bishwanath to predict the future for data centers. The growing trend of cloud adoption will bring in many changes, he answers. Data center consolidation is up ahead. Smaller data center providers will be gobbled up by very large players, unless they have a niche value proposition.

“User IT organizations like ours will have very small inhouse data center footprints, moving most functions on to the cloud as people become more comfortable with it.  The focus will shift from the infrastructure stack to app-level clouds, because user IT organizations will only want to look at the functional part,” he says.

In 3-5 years, we will also see data centers shift out of countries and cluster in specific countries that have cooler climates. Scandinavian countries that have cooler temperatures could become the hot spot for this shift. But for this to happen, network costs have to be much lesser, and statutory requirements will also have to fulfilled.

He explains this with an example. If a data center in Singapore grows considerably, even small improvements in efficiency will drive a huge cost difference. So the data center would be moved to a colder country. And soon others will follow. Countries to which this action will shift will bring in laws making it favorable for data centers to set up shop there. “So instead of having data centers dotted all over the world, they will be consolidated geographically,” says Bishwanath.

Staying Focused and Delivering Value

All this means that the data center as well as user IT organizations like M&M will see a lot of transformation over the coming years. How can one navigate this change?

“I follow a fairly strict review schedule with my teams, set up targets and tend to work with dashboards. I personally work on five or six initiatives on which I do rigorous follow-ups with the team. These are high-value for the organization and will have a large impact. These are also key initiatives to which I will be able to add considerable value. The company could have well over a hundred IT initiatives planned out for the year, but I am not closely involved in most. I do monthly reviews to get a sense of the deliverables and whether they are headed in the right direction,” says Bishwanath.

Sounds like a potentially stress inducing situation, we note. Bishwanath laughs and says, “One can follow Munnabhai’s principle, ‘Tension lene ka nahi, dene ka!’ I don’t get stressed the way people usually do.” Does that mean he has some secret practice that lets him stay calm under pressure? “No, not really. I have always been like this. To unwind, I usually spend some time with my daughter. And I listen to music,” he says.

What advice would he want to share with young IT professionals? “For those in user IT organizations, I would advise them to understand the business and not get glued to the technologies or their implementation. They need to change the company’s orbit using IT. And that will change how the IT department is perceived within the company. Otherwise, you will always be seen as the technical guy. You need to engage with the business, and that’s when people will take your advice. If you want to bring innovations using technology, you will require change in processes, and for that you need to sell the concept to people in business roles. This also means you have to network with the business teams,” he says.

This is indeed great advice from a person whose career has been all about creating business value using IT.