Just when you thought it was safe to move your business into the Cloud, fresh talk this week of making Cloud workloads tradeable commodities. The idea has been around for over a year and someone has finally made a move on it with the Chicago Mercantile Exchange in conjunction with 6fusion announcing this week that they had agreed to “develop a spot exchange for infrastructure as a service.”
The reason for this is that the market has decided that Cloud has matured to a point now whereby it is ready for such as move. Primarily because the difference in service between the large providers is almost minor as to be negligible and the expectation is that over time the service offerings are likely to become identical. Venture Beat described Cloud computing in the same sentence as Soy Beans.
“Soon, you’ll be able to buy and sell contracts for cloud computing resources as easily as you can trade contracts for soybeans and feeder cattle.” – Source
Infrastructure as a Service has become so ubiquitous now, and the margin so low, that we are starting to see the die off of outlying small Cloud providers globally, who simply can’t sustain the ongoing war over pricing. We’re seeing that reflected in the tech stock slump recently as shareholders grow tired with large companies warring in this area still posting negative profits. Infrastructure as a Service in its simplest form, ultimately will be a single-standard offering that most likely Google, Microsoft, and Amazon dominate.
Around the edges of the Infrastructure as a Service die off, Cloud providers are rushing to “anything as a service” recognising that they need unique products, delivered via Cloud, to attract, trap, and retain customers. That’s why we are seeing Software as a Service (Xero) rise and rise and rise. Because people are about over focusing on iron in the Cloud revolution and are now starting to focus heavily on the top of the stack which adds the most value. Look at Cisco as an example, while they will offer IaaS they are also claim a tool belt of over 100 software as a service products now, ditto IBM.
You can expect a slew of products in the tradeable space about to appear. This week’s announcement is one of a couple recently.
“The WAC is the industry’s first patented benchmark to universally measure consumption and capacity of compute, network and storage resources. The WAC makes it possible for buyers to standardize the quantification of IaaS requirements, making it simpler to determine the infrastructure type, quantities, and pricing that best match requirements across heterogeneous suppliers in the marketplace.” – Source
Here’s another from a few months back:
“Berlin, Germany – July 2, 2013 – Zimory, the Berlin-based experts on cloud management software has been chosen as a strategic partner by Deutsche Boerse Group to deliver the cloud technology for its newest venture the ‘Deutsche Boerse Cloud Exchange AG’. Deutsche Boerse Cloud Exchange is creating the world’s first vendor neutral marketplace for cloud computing capacities which will enable IT resources to be traded like securities and energy, electronically, and within seconds. The new company aims to launch initially a cloud computing spot market in the first quarter 2014.” – Source
Forbes hits the nail on the head in their analysis:
“It’s not that the creation of a futures contract produces the profit margin squeeze, it’s rather that the ability to create a futures contract indicates that there’s little left to compete upon except price: that’s what leads to the profits squeeze. This is just great for us consumers of course: but it’s not all that good for the profits that might be made from Microsoft’s Azure, Amazon’s AWS or Google’s Cloud Platform in the future. And we’ve just seen the first effects of such price competition too: Google slashed prices a couple of weeks back and everyone else had to follow suit in order to keep business.” – Source
It’s interesting. It means that we have definitely reached the point globally where Infrastructure a as Service is really like Soy Beans.
This is going to scare customers no doubt, who having just started to really consume Cloud services are now faced with a new risk. Their contract and workloads being traded. It’s far too early to understand what that will do, if anything, to the overall market, but it is likely to slow adoption as companies study what it really means. In the short term, it is another argument for local Cloud providers who won’t be participating in the larger global Cloud commodity market.
Looking to the future, it gets more weird. It is possible that those tradeable workloads are governed by similar Artificial Intelligence that drives the majority of the current trading world. Further, it is possible that we will see a Mega-Cloud develop, controlled by the large players in order to retain those margins.
You see the tradeable aspect of Infrastructure as a Service is likely to keep a cap on any price or margin increases for Infrastructure as a Service, as Forbes notes. So given that those big Cloud providers are burning cash to fight in a Cloud war, how long is it before the seek to raise the price or change the service to increase margin?
How do you do that?
There are different ways that could happen.
You bind your Infrastructure as a Service into a wider service. Think Oracle Cloud. There is no Infrastructure as a Service there is only “something else as a service.” Microsoft is well-positioned, as is Google, to do this. Cisco and IBM are heading in that direction rapidly.
Here’s another way. You work together to create a completely common platform and then you trade the actual workload between each company.
This would mean that the large Cloud providers create and own their own market in which they internally trade workloads between each other where it makes cost sense to do so. For example, if one provider has a few thousand processors idling on the West Coast of North America and another provider is hitting peak load, over subscribed, on the East Coast, then it could lay off that workload to the secondary provider, and vice versa. As capacity in the Cloud grows, this makes sense, because those large providers don’t have to continue necessarily to build physical infrastructure, they can share each others, and agree trading costs between them.
While your gateway to the Cloud might be Microsoft, in the back end your workload could be running on Amazon from time to time, and vice versa.
Why do it this way? Because if the market is going to force you down to a zero cost, you won’t survive, no matter how large you are, so you are better off creating alliances early in order to protect yourself.
Regardless of what happens next, this is going to further complicate what is already a very complex technology. That can only be good, as I’ve said, for the local Cloud provider in the meantime. Companies would be well-advised to go local where they have critical Cloud services and experiment or put low-risk workloads into Global Clouds.
This is new territory in a continuing disruptive technology phase. As The Register, tongue in cheek as usual, put it: “What could go wrong?” Or, as Dilbert put it: