High density in financial services: doing more with less

Written by Sacha Kavanagh, Co-founder and Senior Analyst, Scrutinise Research & Analysis Published 2016-07-12 09:49:41

London is known the world over as a global hub for financial services as well as a hotbed of innovative start-ups in financial technology (fintech). Despite a degree of uncertainty in the wake of the Brexit vote, that is not likely to change any time soon. Established and start-up providers alike have the opportunity to use technology to transform the way companies and individuals raise, lend and transfer money; to increase competitiveness, innovation and efficiency; to differentiate themselves and generate new revenue streams.

Traditional financial service providers are being challenged by young, nimble fintech and insurtech firms exploiting technology to disrupt the industry. Retail banking, investment banking, wealth management, securities trading and insurance are all under attack, often at the most profitable parts of their business.

At the same time, large financial institutions must serve and satisfy the needs of ever more demanding customers who want the same online experience they are accustomed to from the likes of Amazon, whenever they want it and on whatever device they want to use.

And, as if that weren’t enough, they are having to do this in an ever more strict regulatory environment, hampered by legacy IT infrastructure that, even if it isn’t outdated yet, may soon become so as it is overwhelmed by the exponential growth in data.

Many traditional financial services firms still own and operate their own data centres but there is growing acceptance that third parties can provide the quality and security they need, as well as if not better than they can internally.

In contrast, fintech companies have used an opex model from the start, needing the flexibility, scalability and cost-efficiency that third party colocation and cloud providers can supply, thus enabling them to focus resources on their core business from the outset.

Latency has till now been the killer argument for those in financial services seeking to outsource IT operations. Of course latency is still critical for activities such as high frequency trading, but financial institutions do not need to site all their equipment in facilities with ultra low latency. Back-office functions like CRM or HR do not need to be housed in expensive data centres where latency is the be all and end all.

Now, both traditional and newer players alike need their IT infrastructure to be able to cope with ever increasing amounts of data. They need to be able to exploit and monetise that data to develop new products and services, and they also need to be able to deliver services more quickly, both internally and to external customers.

Virtualisation, cloud, high performance computing (HPC) and big data analytics are key to this transformation, but they all need high density computing applications and equipment in order to fulfil their potential. Standard enterprise servers simply aren’t up to the job. Newer, faster, more powerful servers are coming to market all the time, but they are power hungry beasts. As Moore’s Law dictates that computer processing power will double every year, so too must data centre capability be increased.

Simply adding servers to an existing data centre footprint makes no financial sense.

A far more efficient solution is to look for a high power density deployment that will significantly lower the TCO by increasing the per foot computing power of the data centre, thereby maximising utilisation, productivity and efficiency.

As server, storage and network equipment gets smaller and ever more powerful, we’re able to densely pack compute power into a rack to reduce the space and equipment requirement. Although the amount of power per cabinet is higher, far fewer racks are needed so the overall power consumption is reduced. But the higher energy consumption and harder working servers generate substantially more heat, so some form of advanced cooling technology is vital to counteract that.

Not all data centres are created equal, and in older facilities it’s impossible to provide high density within racks without adding cooling and support infrastructure, thus rendering the cost argument null and void. While some data centres have created high density zones, this is not ideal for companies with varying power density requirements.

Clearly, financial institutions of whatever size need future-proofed data centres that can meet their current needs, and provide flexibility and scalability for future growth and performance demands. Modern data centres have been designed and built for high performance and optimisation, and deliver significant advantages over older facilities in terms of power density.

For example, VIRTUS’ LONDON2 data centre in Hayes enables high density (10kW-20kW) and ultra-high density (20kW-40kW) rack options, adjacent to standard racks and without the need for any additional power or cooling infrastructure. For an 80kW deployment, at 40kW/rack this reduces the footprint by 90% and delivers savings of 30-40% on equipment costs and 30% or more on the cost of service, when compared with an industry standard 4kW/rack deployment.

Of course such savings are not only attractive to or achievable by those in financial services. But the sector is one that is at the epicentre of technology innovation, and traditional and start-up firms alike will win, lose or die through technology.

High density is all about doing more with less, the very definition of efficiency. It delivers cost savings and performance improvements, and also ticks energy efficiency boxes for environmental concerns. And with fewer servers, manageability is easier and quicker. Simply put, it’s a no-brainer.