Tuesday, October 17, 2017


Industry relief as ‘death of the data centre’ hailed as not a threat, debate should instead focus on what data centres will be needed in the future



“This ‘death of the data centre’ talk is often seen as ridiculous,” says industry analyst after former Facebook’s VP comments on the development of new chipsets send stocks down across all the major US data centre REITs.

The data centre industry was hit by an earthquake after comments made by an influential venture capitalist and hedge-fund owner sent the stocks of all American data centre REITs down with the premise that future processor technology will result in the end of the data centre provider industry.

The industry was quick to react with many taking to the usual social media channels including LinkedIn, Twitter and Facebook, some agreeing and some disagreeing.

The comments were made by none other than 41-year-old Chamath Palihapitiya, founder and CEO of the VC firm Social Capital, who has also served as Facebook’s VP from 2007 and 2011.

According to Seeking Alpha, citing Bloomberg, Palihapitiya was commenting on the news that Google may have developed its own chip that can run 50% of its computing on 10% of the silicon.

He said: “We can literally take a rack of servers that can basically replace seven or eight data centres and park it, drive it in an RV and park it beside a data centre.

“Plug it into some air conditioning and power and it will take those data centres out of business.”

The market view sent stocks down, and according to Seeking Alpha, this was the scenario after the comments were made public: Digital Realty Trust (NYSE:DLR) was off 3.6%, while DuPont Fabros (NYSE:DFT) was 3.4% lower. CoreSite Realty (NYSE:COR) was down 3.5%, and QTS Realty Trust (NYSE:QTS) was down 3%. CyrusOne (NASDAQ:CONE) ias 2.4% lower and Equinix (NASDAQ:EQIX) was down 2.1%.

Palihapitiya would later react to the market’s take on his view: “It is not going to happen overnight, but investing is not about overnight, and if you can see the train wrecks coming you can take advantage of them.”

Reacting to the panic generated, analysts have now come out to explain what the industry really needs to worry about, in an attempt to ease the markets.

Steve Wallage, MD of BroadGroup Consulting, said: “To old hands in the industry, this ‘death of the data centre’ talk is often seen as ridiculous in a world of soaring demand for data and given the lack of churn in the industry, importance of reliability and security, and such revolutionary technology claims historically failing to make much impact.

“Some of these arguments have validity and the development of data centres has tended to be evolutionary and based on proven technology, rather than revolutionary, but the industry needs to avoid complacency and understand better some of the new technology on the horizon.”

However, while the analyst plays down the fear of the disappearance of the data centre provider industry, he warned that market players have instead to discuss more on what sort of data centres will be needed in the future and where they need to be located.

Wallage said: “Some co-location players remain stuck in the mindset that Tier III, downtown, ‘traditional data centre hubs’ will meet all demand. Furthermore, that their 10-15 (or even 20) year old sites can continued to be retro-fitted and remain attractive.

“What exactly that future demand will look like is difficult to predict – however, one example of the present challenge is the focus of many telcos on edge computing compared with the ‘lip service’ paid to it by many co-lo players.

“We are already seeing newer players offering far more flexibility in their data centre designs and builds, optimising solutions for different verticals and requirements, and realizing that many data centre requirements need a mix of proximity, more cost-effective locations and networked sites.

“So data centres will continue to evolve and those who adapt to market demands through whichever technologies are adopted, will remain among the longer term winners.”

Wallage’s comments mirror the discussion that has been happening at many data centre and cloud events around the world and the views of several industry leaders.

The need to innovate on a business roadmap level, rather than on the technology stack, seems to be the key for survival of data centre providers.

Edge computing, as mentioned by Wallage, is indeed a good example of a new business stream, whose technology is already available, but deployments are sometimes faced with cultural barriers in boardrooms.

Just like the ongoing debate surrounding the end of the mainframe, data centres providers will most unlikely be killed off based on the current state of adoption, growth and ever souring demand.

However, just as the business models used in the past do not function today, today’s strategies will become obsolete as market consolidation and grow occurs.

As Data Economy has previously put it, data centres are tomorrow’s oil refineries, and this is a market that – independently of being colocation, wholesale, retail, private or public cloud – will not disappear, but transform.