AWS Chronicles

So my question is when does Amazon.com split its retail operations from its AWS platform business. I’d love to see these priced separately. Actually, truth be told, I suggest Amazon.com is actually three businesses:

• the AWS computing platform
• the Amazon retail and logistics platform
• the Amazon.xxx online store(s)

At the risk of being accused of adding only ‘thin’ value, I would suggest that these three businesses run and capitalized individually would be worth more than Amazon’s current $60bn market cap. Indeed, Amazon.com is a perfect example of a firm that is natively adapted to the new optimal ‘industrial stack’:

Earlier this year I suggested that AWS in particular could well be the totemic representative technology that inaugurates the sixth techno-economic paradigm:

Just as Intel’s 4004 microprocessor was the catalyst for a wave of creative destruction in the 70s and 80s, will AWS prove the same for the 00s and 10s? Probably. We’re seeing it already. And it’s going to disrupt the hell out of the mastodons of industry across most sectors of the economy. Why? Because their cultures and leaders are entirely ill-equipped to face such a fundamental paradigm shift. They know how to play by the old rules. The strategic competitive advantages they built up over decades risk suddenly – poof! – to become obsolete.

And then a couple of weeks ago, Amazon announces spot instances on EC2. Amazon’s CTO Werner Vogel explains:

The central concept in this new option is that of the Spot Price, which we determine based on current supply and demand and will fluctuate periodically. If the maximum price a customer has bid exceeds the current Spot Price then their instances will be run, priced at the current Spot Price. If the Spot Price rises above the customer’s bid, their instances will be terminated and restarted (if the customer wants it restarted at all) when the Spot Price falls below the customer’s bid. This gives customers exact control over the maximum cost they are incurring for their workloads, and often will provide them with substantial savings. It is important to note that customers will pay only the existing Spot Price; the maximum price just specifies how much a customer is willing to pay for capacity as the Spot Price changes.

Spot Instances are ideal for Amazon EC2 customers who have workloads that are flexible as to when its tasks are run. These can be incidental tasks, such as the analysis of a particular dataset, or tasks where the amount of work to be done is almost never finished, such as media conversion from a Hollywood’s studio’s movie vault, or web crawling for a search indexing company. For most of these tasks their completion is not time critical and as such they are ideal targets for additional cost savings.

Before I go any further, let’s just say it’s pretty exciting to see vision become reality even if in this case I’m only a distant spectator. Markets in anything. Digital markets. Themes that go back to the founding mission statement of the Park Paradigm:

(December 2005) The technology of the digital age is driving an unprecedented explosion in the ability to create markets in anything. Trade anything. Not just physical goods. Not just financial instruments. But ideas. Events. Outcomes.
The emergence of these kinds of markets will – over time – impact how we view and interact with the world in all aspects of our personal and professional lives. They will fundamentally alter the current world economic and social paradigm.

Chris Swan calls them virtual resource markets and correctly points out that, at least for now, the market is “closed” – ie users cannot trade their capacity amongst themselves, however I suspect that it is just a matter of time before such a market is organized. But what would be even more useful (and exciting) than a closed market on Amazon EC2 resources, would be an open marketplace for on-demand spot computing resources. ie A marketplace which is agnostic as to where the compute resource comes from, so long as it is a robust and more or less uniform resource.* However for this to be useful for the end consumers of this computing commodity, the ability to switch automatically and seamlessly from one cloud computing source to another based on price and/or availability would be crucial. Indeed this would be the key value driver for anyone hoping to operate a compute resource exchange. Sure the price discovery and transaction mechanisms would be necessary but these are relatively trivial to build and hard (in isolation) to monetize. The real value creator for any exchange (just ask the CME) lies in clearing and settlements. (For the non-financial amongst my readers this is the back-end of the trade, fulfillment essentially.)**

James Urquhart makes this point strongly in his review of spot instances:

Note, however, that this feature is not market-based pricing. Amazon determines the spot price and can raise that price enough to gain back capacity at will, at no real cost to itself. There is no competition. There is no commoditization. There is just consumption of what is not being used.

The truth is, real commoditization of infrastructure services–or any other cloud service, for that matter–isn’t in the best interest of Amazon or any other service provider.

Regardless, commoditization can’t happen without open standards that allow easy portability and interoperability of data and code, as well as security, control, service-level assurance and compliance systems. Those standards are coming, but it is impossible to predict when they will arrive. I only hope Amazon embraces them when they do.

I’m not sure I agree with his view however that commoditization isn’t in the best interest of Amazon. The underlying asset is ultimately relatively undifferentiated (a compute cycle is a compute cycle is a compute cycle) which is in fact the definition of a commodity. If you are a provider of a commodity – unless you can maintain a monopoly or a cartel – it is in your interest to create as big and vibrant a marketplace as possible. Supply creating demand. And particularly if you fancy yourself the most efficient, large scale producer of said commodity (as I’m sure Amazon does), all the more reason to want a big, liquid market of consumers. It is the exchange and clearer that want to create lock-in, not the producers. To be fair, for the moment AWS is both and indeed this is the point James is making I think, but I would be surprised if they had the intention (hubris?) to think this is anything but a transitionary arrangement.

Of course, as a traded market in this critical 21st century resource develops over the next decade and beyond, the business opportunities abound. Better yet, many of them are well known and can quickly be adapted (from other asset markets) to apply to the compute resource market. It’s not a business yet, but it only took a few hours before the first ticker tapes (here also) began to appear for EC2 pricing:

An entire ecosystem will surely emerge – exchanges, prime brokers, risk management derivatives, algorithmic trading… I’m sure there will also be some interesting second-order opportunities. Linking spot computing prices with spot electricity prices. Selling green compute cycles (ie powered by renewable energy sources only.) Allowing anyone to sell compute cycles into the grid (think SETI@home meets micro-generation). The mind races.

Welcome to the sixth paradigm.

** I can’t help but wondering if the amazing technology developed by our portfolio company CohesiveFT couldn’t be adapted or re-purposed to form the core fulfillment engine of a compute resource exchange. The fact that they are Chicago based and their CEO/Founder is ex-O’Connor makes me wonder even more!

Related articles by Zemanta

• Amazon expands Web-based business services to Asia (seattletimes.nwsource.com)
• How the cloud could conquer the world in 2010 (theregister.co.uk)
• Amazon Virtual Private Cloud Opens Up (aws.typepad.com)
• Wonderland. (parkparadigm.com)
• Rulers of the Cloud: Will Amazon’s Computing Fabric Become a New Economy? (readwriteweb.com)
• Could A Cloud Computing Exchange Work? (datacenterknowledge.com)

Tweet This Post

Permalink||Comments|Filed under 'Business, Sixth Paradigm, Tools, Web X.0, markets'