The implacable march of Amazon Web Services toward ultimate public cloud domination has been relentless, from its inception in 2006 with a single service (S3 Storage) to the behemoth it has become today. It seems this minnow has become the biggest fish in the pond. But is it unstoppable? Has it won the public cloud wars?
That is still up for grabs. There are still some major players out there that could move. IBM Softlayer? Unlikely. Azure? Definitely. Oracle Cloud? Maybe. You will notice that Cisco is not there. This is because it has just pulled out of that market. Yes, Cisco is closing its $1 billion investment in its Intercloud product in March 2017. According to a Cisco source, it will be “moving current workloads into alternate infrastructure, including, in some cases, public clouds.” Cisco will follow in the veritable footsteps of former cloud illuminatus Rackspace, which is now effectively an AWS and Azure provider. Although VMware still has vCloud Air, now that it is on AWS play, it is effectively another AWS reseller/partner.
How this came to pass is a modern-day lesson in business ostrich syndrome, similar to what has become known as the “Kodak blindness.” Another company that displayed the same blindness is Nokia.
Cisco is just the latest in a long line of legacy vendors that never understood the real drive to the cloud and as such could not, or rather would not, invest in the transformation within their own businesses that such a disruptive technology change requires.
Cisco is heavily reliant on the CapEx-driven upgrade cycle, which regularly brought large amounts of cash into its business. Its customers have grown weary of coming up with large amounts of cash every four to five years. They are tired of having to purchase more hardware than they need in order to get the necessary discounts and to cover the growth they hope to see in the following four to five years. This ploy has never worked out well. The demand for hardware is either underestimated, drawing the wrath of the board for having to buy more gear later, or worse yet, overestimated, with hardware sitting underutilized through its end of life.
Suffice it to say that businesses were ripe for a change, and the public cloud offered this. They could just purchase what they required; they could grow as and when needed. Further, they could contract if necessary. More importantly, at a monthly charge, this came from their OpEx budgets. This obviated the need to find large amounts of cash every couple of years, usually having to be raised by finance. Boards and businesses love this model, and AWS, as an early mover, has reaped the benefits.
VMware, Cisco, and others attempting to move into this market have done so in a halfhearted manner, without the necessary investment in money or board backing. True, a $1 billion investment seems a large number, but consider this: Amazon invested over $12 billion in technology and services in fiscal year 2015 alone. That number has risen almost sevenfold since 2010.
Has Amazon Won?
On the face of it, it would appear so. However, Oracle, once a serious cloud denier, has entered the market very aggressively in the last year or so, moving to the third largest public cloud by revenue and overtaking Google. It has a big advantage in that it owns its own hardware stack and software stacks, and over the last three years has been quietly reengineering its products for SaaS and PaaS migration. Further, it has been quietly bagging some serious cloud engineers from competitors. Oracle has woken up and smelled the coffee, and it is serious about grabbing as big a cup as possible.
The second biggest public cloud provider is Microsoft. It is again late to the party, but due to its massive presence in the enterprise, it has rapidly gained market share. Azure is mature in terms of services offered, and both Azure and AWS offer storage, compute, etc. However, both offer Hadoop support: Elastic MapReduce (AWS) and HDInsight (Azure). Azure offers more granular payment, billing by the minute rather than the hour, but the core thing that both Oracle and Microsoft offer—which to me is the killer app—is the ability to be truly hybrid. With Oracle Cloud Machine and Azure Stack, both offer the ability to have the best of both worlds: keep your crown jewels within your bastioned towers and move your cattle and grain out to the public cloud. This will be the true battlefield. Oracle and Microsoft have taken significant market share, moving to second and third place by revenue in a short time. While it is true that their combined revenue is still six times less than that of AWS, I firmly believe that, due to their ability to support a hybrid environment, they will find it easier to gain clients.