Oleksii - stock.adobe.com

How DBS is pushing the limits of IT efficiency

Besides shrinking its datacentre footprint by 75%, Singapore’s DBS Bank is running twice as many virtual machines on a single socket and decoupling applications from its infrastructure stack

In a recent Harvard Business Review study, Singapore’s DBS Bank joined the ranks of Netflix, Amazon and Alibaba as the world’s top 20 companies that have made the most successful business transformations in the past decade.

That DBS was the only Singapore company in the list was notable, but more laudable was that the bank only started transforming its infrastructure around four to five years ago, according to its newly minted group CIO, Jimmy Ng.

Speaking at VMware’s vForum 2019 in Singapore last week, Ng said the bank had two options as it was planning to transform its infrastructure – tear down existing datacentres and move its systems to the public cloud, or build a virtual private cloud (VPC) which it eventually did.

But as not all the bank’s applications were cloud-ready, it started building new cloud-native applications while refactoring its legacy applications for the cloud, said Ng.

“Today, we’re 99% virtualised, and that’s only the first step of modernising our applications,” Ng said, adding that DBS had also applied “great engineering” in the way it set up its VPC, which can host twice the number of virtual machines per socket than what is commonly practised in the industry.

This engineering feat is also seen in DBS’s efforts to reduce its datacentre footprint. In 2017, it worked with Equinix to transform one of its traditional datacentres into a smaller cloud-optimised one with a lower operating cost.

The initiative enabled DBS – Southeast Asia’s largest bank by market capitalisation – to move its main datacentre to a facility a quarter of the size of its existing datacentre.

“Today, we’re 99% virtualised, and that’s only the first step of modernising our applications”
Jimmy Ng, DBS Bank

Ng said even with a smaller datacentre, DBS has 10 times more capacity than before to run existing workloads. “The amount of work and engineering involved has reaped a lot of benefits for us,” he said.

With its infrastructure running near peak efficiency, Ng said the bank had to strive for the next level of efficiency at the application level – by decoupling applications from the infrastructure stack by using containers.

Today, about 10-15% of the bank’s workloads are running on VMware’s vSphere Integrated Containers, Ng said, adding that “containers are going to be next space that we’re going into”.

An earlier interview with Ng’s predecessor, David Gledhill, revealed that DBS is also using Red Hat OpenShift and Pivotal Cloud Foundry as its container application platforms. Gledhill said DBS chose the two platforms because it “didn’t want to bet on just one horse”.

“Frankly, we still don’t know who will be the winner of that race, so maybe we will stay with two platforms which still have a lot of components that we want to use,” he said.

DBS’s efforts to improve IT efficiency and drive innovation are exactly what CIOs in Asia-Pacific (APAC) will need to do as they become directly responsible for business outcomes.

According to a VMware survey of 200 CIOs in the region, 64% of CIOs believe they will become key decision-makers for corporate strategy, with nearly 60% of CIOs in the region expecting to head a profit centre and 56% expecting to become CEOs by 2025.

Read more about digital transformation at DBS Bank

Looking ahead to 2025, CIOs are expecting a strategic pivot in their roles. They are not only technology implementers but also architects of new business models, overseeing how their organisations build, run, manage, connect and protect workloads in response to the rapidly developing app economy.

APAC CIOs are also expected to embrace cutting-edge, data-driven technologies to stay competitive and unlock new growth opportunities for their businesses.

The top five future technologies that APAC CIOs in the study are betting on are machine learning (63%), the internet of things (61%), artificial intelligence (60%), edge computing (57%) and blockchain (51%).

Read more on IT efficiency and sustainability

Data Center
Data Management