Pulling IBM’s cloud investments together

The announcement by IBM of a $1.2bn investment in serious expansion of its global cloud services footprint has prompted several comments along the lines of it being a last-ditch attempt to recover the company’s fortunes. Yes, the company’s hardware sales – especially server sales - are dropping, as they are for the majority of the larger vendors, but hardware sales have been a minority share of IBM’s revenue stream for a while.

It seems clear that this is no more the case of staving off corporate implosion than any business decision is for any company, any day of the week. The key part of the announcement is to be found elsewhere, in the words `cloud services.’

The company is investing in 15 additional datacentres around the world this year, in locations such as China, Washington D.C., Hong Kong, London, and India. Centres in the Middle East and Africa are scheduled to follow in 2015. These are in addition to the existing global footprint of 13 datacentres from SoftLayer and 12 from IBM.

The additions would be a bold – and extremely risky – investment if they were purely to add capacity for resale as a hosting service of any kind. That would put the company head-to-head with the likes of Amazon, Rackspace and the rest where the tendency is to compete on price – and with the costs of a $3.2bn total investment to service, that would be a fool’s errand indeed.

So the important point here is that this huge investment is being made at a time when the cloud is just starting to really move centre stage: IBM’s timing might just be right, in practice.

Also, the company has a huge army of customers that are also business partners through being service providers in their own right - from small ISVs through to major MSPs and the like. It can therefore not only provide the compute, network and storage resources, but also the services and support that could make it the service aggregators’ aggregator.

And that could happen at a time when the long-foreseen potential of service aggregation – where companies pull together the applications, tools, support and consultancy services needed to address the full range of business needs of a market sector, providing them with packages of services that meet those needs – is starting to take off.

Much has been made in press comment about IBM’s $2bn acquisition of SoftLayer last year – but that is only a support development where not even the addition of 13 datacentres is that important. Instead, it was the software and management tools for dealing with high volume cloud storage services and the like that was the more important prize.

But the most important component of IBM’s recent development is almost certainly the $1.2bn investment in The Watson Group, announcedtwo weeks ago. The initial applications from the Group are big data related and delivered by cloud. These are Watson Discovery Advisor, for accelerating research and development processes, Watson Analytics for visualising big data insights, and Watson Explorer, which helps users across an enterprise uncover and share data-driven insights more easily and move into big data initiatives faster.

These three could be a big money spinner for the company because they have the potential to be the basis of a wide range of business services, quite possibly attached to service aggregation offerings, that can help businesses make sense – and money – out of big data. And it will be services like that – rather than having lots of storage capacity available – that will generate the real revenues from the cloud.

Of course, once a business got into the idea of exploiting such services it would be a safer, more logical step for them to also consider utilising IBM’s cloud hosting resources, either directly or through one of its MSP and Business Software Solutions partners, many of which would also be in a position to aggregate additional services into the overall service mix.

By some estimates, the global cloud market is set to grow to $200 billion by 2020, driven largely by businesses and government agencies deploying cloud services to develop products, market them and manage their supply chains in transforming their business practices.

An interesting side-issue to the IBM announcement is how it might impact the smaller Cloud Service Providers (CSPs) and what they might do about it. This also demonstrates the many facets that need to be considered when it comes to building or using cloud services.  

Sven Freudenfeld, Business Development Director, Telecom-Cloud at Kontron, suggests that the CSPs can respond by setting new standards for cloud infrastructure in terms of cost savings and power efficiencies, while delivering new advances in media server technology. Other CSPs can add value by implementing solutions that provide affordable, fast, energy efficient processing infrastructures, optimised to support the cost effective delivery of business applications and content services, he contends.

“The critical element for the CSP in today’s data hungry environment is system, power, and cloud management,” he said. “Many datacentres are still reliant on power hungry and expensive legacy equipment that isn’t capable of supporting the delivery of video or other media rich content. These systems are not as efficient as the high-density modular designs that are now available. Cloud service providers can now adopt media server technology that harnesses 4th generation Intel Core technology and a scalable platform that provides tremendous cost versus performance benefits.”

There is certainly much truth in that suggestion, especially when it comes to power consumption. But it is also fair to observe that we are still in the early days developing the applications and management software that provide such services, and much of it could arguably be better designed to make far more efficient use of the resources that are available. And it is the software that will provide the services, with hardware in most cases playing only a supporting role.