Any new technology presents users -- and early adopters in particular -- with an uneasy combination of extravagant claims and immature execution. In many cases, the users with the most compelling business cases will take the risks and blaze a trail for others. When they do, they leave signposts along the way in the form of the impact of their adoption on the nature and prices of services. That's the way it has always been, and that's the way it is with the cloud. It's time we read three of those signs to see where the cloud revolution is heading.
1. IaaS price cuts. The most important market signal of the last six months is the continued decline in pricing for infrastructure as a service (IaaS). This decline is an indication that cloud providers are seeking more business by cutting prices. That's good news for the cloud consumer. However, price commoditization will make it much harder for new innovative companies to enter or survive in the IaaS market.
How will the constant reductions in IaaS pricing be viewed -- as good or as bad news?
2. IBM's server move. Another critical sign is IBM's decision to sell its x86 server business to Lenovo. It's not that IBM thinks that x86 systems have no future, but that the prices and profit margins for x86s will be under continual pressure -- not the sort of thing a big technology company wants to see. From a cloud perspective, commodity servers are a good thing because they'd cut the costs for cloud providers, and they'd also cut the costs for users who are comparing cloud costs with those of internal IT.
How will this critical development be viewed from a cloud perspective?
3. PaaS and SaaS growth. Finally, we're seeing more platform as a service (PaaS) and software as a service (SaaS) growth, as well as a growing market for "platform services," or Web services to enhance basic IaaS offerings.
Why are users moving up the cloud stack? And why are providers encouraging it by focusing more on higher-layer services, those above IaaS?
These three market signs have seemingly little in common. But taken together, they signal a radical shift in the paradigm driving cloud adoption, and that's a signal of a different future for the cloud.
IaaS price cuts are a signal that the cloud giants recognize adoption will depend entirely on cloud price. The case for hosting a current app in the cloud is created by the difference between its in-house and cloud costs. Cloud cost reductions are aimed at getting more qualifying applications. This means that the adoption rate of the cloud would likely decelerate, at least in the near term, without a boost in cloud savings. The simple "virtual hosting" model of the cloud is probably already running out of steam.
All this means that cloud providers won't become better businesses because lower prices create more users; they'll become better businesses by selling basic IaaS users other cloud services on top of IaaS.
It's also running out of profits. The cloud providers lose revenue on every existing customer when they drop prices. Contrary to what's been written, the major cloud providers have little further economies of scale to gain.
IBM's decision to leave the x86 server market shows that the hardware there is already commoditized, so few more cost reductions in basic cloud infrastructure can be expected. All this means that cloud providers won't become better businesses because lower prices create more users; they'll become better businesses by selling basic IaaS users other cloud services on top of IaaS. The lower prices are a signal that IaaS might actually become a loss leader to get users into the cloud store to buy virtual "milk," then buy virtual "cookies" while they're there.
Moving above basic IaaS
Two things happen when you move "above" basic IaaS services in the cloud. First, your new services displace more hardware and software cost on the user side. Second, users can then justify higher prices from a cloud provider without killing cloud's incentive. IaaS displaces only hardware cost; PaaS displaces hardware, OS and middleware costs; and SaaS displaces all application costs.
The most important impact of a move from IaaS is allowing cloud providers to begin to support native cloud application development. Amazon's new Web services are clearly not designed to induce people to migrate their existing content-caching or Web-accelerating applications to the cloud; they have no such applications. It's these applications, things that can be done well only on the cloud. They will evolve from the limited "hosted server consolidation" applications of today and drive the cloud's future.
Amazon, Google, Microsoft and other cloud providers need a customer base so they can sell their cloud-specific services. Price reductions for IaaS will keep that base, and opportunities to upsell into the emerging cloud-specific service market will grow.
The more higher-layer components there are to cloud services and the more cloud-specific that applications become, the less the cloud of the future will look like the cloud of today. The notion of a "hybrid cloud" will almost disappear as agile components move freely across hosting options using these advanced orchestration concepts. Every cloud will be potentially a hybrid, so users and providers will rely on deployment and management tools that converge on a common model.
The signs are clear: The cloud is not a different hosting option for existing applications; it's a different architecture for application development. Adding the cloud hosting dimension to the tools available to developers will change forever how we write business and even entertainment applications and services. For those who relish revolution, the best news of all is that the price wars and service trends we see today are a clear sign that the cloud providers themselves think a cloud revolution is just around the corner.
About the author:
Tom Nolle is president of CIMI Corp., a strategic consulting firm specializing in telecommunications and data communications since 1982.