(originally published march 2014)
IBM has officially dedicated over 1.2 billion USD in additional investments to its new cloud data center infrastructure. That makes roughly 80-100 million USD for each of the 15 newly planned data center locations (15 new DC`s adding to 13 Softlayer and 15 IBM data centers) that will cover all geographic regions. Although IBM has tremendous in-house expertise when it comes to building and operating data center facilities, the new cloud data centers will follow the blueprint of Softlayer! Why that?
As a typical hosting provider Softlayer does not only command a highly scalable data center architecture with over 80,000 running servers. It is foremost the highly cost-efficient data center operations and broad set of services that makes it attractive for IBM to switch platforms from its own IBM SmartCloud to the Softlayer platform. What Softlayer also brings to the table is a substantial revenue share. Crisp Research estimates the annual revenue for 2013 around 520 million USD. Given the new investments and the IBM brand as a backing for enterprise deals – the 2014 revenue can easily be around 600 million up to 1 billion USD.
Within a positive market scenario the new IBM/Softlayer business unit is likely to surpass 4 billion USD in revenue within the next 5 years and thereby contributing up to 10% of IBM GTS revenue by the end of 2018.
Softlayer to become IBM`s prime cloud delivery platform
But this is only possible if IBM will stick to its current roadmap and make Softlayer the prime cloud delivery platform for its new SaaS and PaaS-offerings. As of today IBM plans to make most of its software portfolio accessible as cloud services – within shared, hybrid and private cloud environments. For example, the CloudFoundry-based IBM Bluemix-offerings (PaaS) is already being hosted on the Softlayer platform. The same will happen to new collaboration services (e.g. Connect), infrastructure management solutions and even “Watson”-based analytics services.
And there is even more evidence that IBM is getting down to business with Softlayer. Relatively unnoticed IBM has migrated hundreds of customers from its SmartCloud to the Softlayer platform over the course of the last 5 months. It could be stated that this was one of the largest cloud migrations (vendor-to-vendor) - by the number of clients. But not in terms of capacity. Most of the workloads were “under the radar” applications.
The IBM SmartCloud has been never been more than a stepchild in IBM´s overall strategy. The platform was not capable of handling thousands of agile and scalable web workloads in a competitive operations model. Thereby the IBM senior management team has come to the right decision – let Softlayer grow – shut SmartCloud down.
Bare metal cloud hosting– the secret sauce in the IaaS market?
Over the last 5 years the cloud market has gone a long way in terms of technical maturity and overall adoption. And what vendors and CIOs both learned alike is that there is a clear demand for different types of cloud infrastructure services. One-size-fits all virtual machines are not the right solution for all customers. Workloads have highly differentiated characteristics. Especially for advanced workloads “dedicated performance” instead of mellow SLAs can help the customers to sleep better. Business cases vary in terms of flexibility and cost. As more and more workloads are migrated towards the cloud “cost prediction” becomes key to numerous CIOs and CTOs. To cover baseline infrastructure capacity reserved resources (e.g. bare metal or dedicated VMs) help to save budget. These savings can be allocated to innovation projects or to cover “peak” capacity.
The market for infrastructure-as-a-service (IaaS) is still under development and covers a large spectrum of providers and delivery models. Ranging from the large public cloud platform providers AWS, Google and Microsoft Azure who offer mainly standardized services to the classical managed service providers with their private cloud offerings. That gap is being bridged by a small number of hosting firms that have understood to transform their portfolio in direction of cloud. Softlayer is among the group of Host Europe Group, OVH, Rackspace et al. that combine agility, dedicated performance and cost predictability in the right balance together with solid support operations. Looking at the growth numbers delivered by this peer group over the last years we can reason that the “cloud hosting” model has a huge fan base in the market. This is mainly true for SMB customers, eCommerce firms and part of the startup world.
Remaining questions and outlook
Softlayer has a lot to offer for IBM. And IBM seems to appreciate the new capabilities and innovations. But some strategic questions are not answered yet. For example: How does the Softlayer hosting business fits with the IBM high-margin model on the long run? Even when IBM is able to drag enterprise customers (who are willing to pay more for dedicated compute power then for virtual compute power) onto its new platform the IaaS business will not have the same margins like the Mainframe business.
One next question relates to IBM´s plan to migrate its entire software portfolio to the cloud (“SaaS First”). The ambitious growth plan for Softlayer will only pay off if the migration and go-to-market for the different software solutions is successful and well communicated. Currently it is not quite clear which of the hundreds of software solutions are able to run in the cloud and which of them will be migrated first.
But the most important question is: Will IBM stick to its plan? From the perspective of Crisp Research there is no way out. After the current announcements customers are betting on IBM Softlayer. SmartCloud is not a trusted platform any longer and will go to the dead pool. IBM´s current activities look very promising. But competition is fierce. Rich Miller estimated a whopping 7.3 billion USD that were spend by Google in 2013 to expand its data center and network capacity (up from 3.2 billion USD in 2012).
With Softlayer as its new cloud quarterback IBM is back in the game. The next 12-18 month will show, if AWS, Microsoft and Google will get serious competition in the cloud capacity game where “economies of scale” dictate the rules.