Back during the dot.com boom years, existing telcos and dozens of new network operators, especially in western Europe and North America, laid vast amounts of fiber optic networks in anticipation of rapidly rising Internet usage and traffic. When the expected volumes of Internet usage failed to materialize, they did not turn on or “light up” most (some estimate 80% and even 90% on many routes) of this fiber network capacity. This unused capacity was called “dark fiber,” and it has only been in recent years that this dark fiber has been put to use.
I am seeing early signs of something similar in the build-out of infrastructure-as-a-service (IaaS) cloud offerings. Of course, the data centers of servers, storage devices, and networks that IaaS vendors need can scale up in a more linear fashion (add another rack of blade servers as needed to support an new client) than the all-or-nothing build-out of fiber optic networks, so the magnitude of “dark cloud” will never reach the magnitude of “dark fiber.” Nonetheless, if current trends continue and accelerate, there is a real potential for IaaS wannabes creating a glut of “dark cloud” capacity that exceeds actual demand, with resulting downward pressure on prices and shakeouts of unsuccessful IaaS providers.
One set of evidence comes from the wide range of vendors that have publicly announced cloud computing strategies, especially on IaaS offerings. In addition to the current IaaS vendors like Amazon.com, Google, Rackspace, and Savvis (being acquired by CenturyLink, which also recently bought major U.S. telco Qwest), hardware vendors like EMC, HP, and IBM, and IT outsourcing vendors like Accenture, Capgemini, Infosys, TCS, and Wipro, along with virtually every major telco, have announced their own cloud computing strategies. So there are easily three or more dozen — and potentially many more — major players who have or plan to have IaaS offerings.
Of course, strategy announcements do not necessarily translate into investments in capacity. Indeed, my colleague Brownlee Thomas, who covers the telecoms sector, told me that the telcos she talks with have been fairly cautious about throwing a lot of money at building out cloud capacity, mostly doing so when they have major clients asking for it, and specifically for “private” cloud offerings. (However, as my colleague, James Staten, points out in his market overview of private cloud solutions, it’s the leading hardware vendors – not the telcos – that are taking command of the private cloud, so telcos may be pushed into the less predictable public cloud space. )
But here’s where the other piece of evidence comes in. Looking at the calendar Q1 2011 financial results for leading hardware vendors, it is striking how strong sales of server and storage equipment were.
- At this stage of the hardware replacement cycle and with virtualization dampening demand, we expected server revenues to post growth rates of around 5%. But IBM reported hardware revenue growth of 22%, Dell had server growth of 10%, and HP had server growth of 8%. Cisco, in conversations with Forrester analysts, said that its data center products – especially its UCS blade server products – were up 38%.
- HP and IBM reported increases in their sales to other parts of their organization, presumably in support of their build-out of their own IaaS offerings. IBM’s hardware sales to other parts of IBM rose by $71 million, from $173 million in Q1 2010 to $244 million in Q1 2011. The line in HP’s income statement for “eliminations of intersegment net revenue and other” in HP’s most recent quarter rose $126 million, from $656 million in 2010 to $782 million in 2011.
- Storage vendors like EMC and NetApp that sell the mid-range storage gear most likely to be used in cloud deployments reported strong storage product revenue growth of 18% for EMC and 27% for NetApp. Indeed, EMC, in its Q1 2011 10-Q filing, reported that “The overall growth in [storage] product revenue was due to a continued higher demand for our IT infrastructure offerings to address the storage needs for continued information growth, particularly as customers continue to build out their own data centers to develop and support their private or public cloud infrastructures [emphasis added].
So, there is clear evidence of cloud computing wannabes buying significant quantities of hardware gear to support their ambitions. Back-of-the envelope calculations suggest that cloud computing wannabes invested around $1 billion in equipment in Q1 2011, with the potential for investing $6 to $10 billion for 2011 as a whole.
Now, against that both potential and actual increase in the supply of IaaS offerings, look at the potential demand. In their April 21, 2011 report, “Sizing The Cloud — Understanding And Quantifying The Future Of Cloud Computing,” Forrester analysts Holger Kisker and Stefan Ried projected that public-cloud IaaS revenues would rise from $1 billion in 2010 to $3 billion in 2011 and then $5 billion in 2012, then level off at around $6 billion. Virtual private cloud revenues for dynamic infrastructure services will rise from $0.8 billion in 2010 to $1.4 billion in 2011, $2.5 billion in 2012, and $4.5 billion in 2013. These are impressive growth numbers, with even larger growth projected for virtual private clouds beyond 2013. However, the combined total of $3.4 billion in public and virtual private cloud revenues in 2011 and $7.5 billion in 2012 will probably be less than the investment by IaaS providers in data centers and equipment to support these offerings, leaving many vendors operating at a loss.
Could the Q1 buying of hardware be a temporary blip that will fade away? Certainly possible. Could the adoption of IaaS be even stronger than Forrester projects? Again, certainly possible. Will the IaaS vendors be measured and cautious in their investments in support of their IaaS offerings, only adding capacity when they have actual paying customers? Hopefully.
But remember the line of former Citibank CEO Charles Prince about why his firm was continuing to lend to private equity buy-outs and other leveraged financial instruments during the financial market boom: “As long as the music is playing, you’ve got to get up and dance” (Financial Times, July 9, 2007). That logic drove financial services firms to continue playing in the over-heighted debt markets, despite private doubts, because investors punished firms who weren’t reaping the profits from being in that market. A similar story could happen in the cloud computing market. No major tech vendor or telecom carrier feels it can stand aside from the cloud computing boom, because there is real growth there. But if everyone follows through with their plans, there is likely to be more capacity than demand. So I think the prospect for the emergence of dark cloud is a real one.