Amazon Web Services has cut prices on “dedicated” EC2 compute instances by up to 80 percent (!) in some cases. That’s a big price chop even by AWS (s amzn) standards.
An instance is a unit of compute power that customers rent by the hour and is a key base unit of AWS or other cloud infrastructure. Increasingly customers are renting compute power in AWS instead of buying their own physical servers.
These dedicated instances run on single-tenant hardware that is allocated to one customer account while other instance types run on shared infrastructure. That’s an important implementation option for businesses or government agencies — in financial services, health care, insurance and other regulated entities — that are not supposed to put sensitive workloads on shared hardware.
The move comes at a time when Amazon contends with cloud competitors who maintain that they, unlike AWS, offer private cloud capabilities better suited for regulation-constrained industries. And AWS, which leads the league in public cloud infrastructure also faces several rivals — Google(s goog) Compute Engine, Microsoft (s msft) Windows Azure, ProfitBricks and CloudSigma — which offer sub-hour incremental pricing on their instances.
The stakes are high. Gartner expects the overall market for public cloud services to grow 18.5 percent this year to $130 billion from $111 billion in 2012. All of the aforementioned companies along with Dell(s dell), IBM (s ibm), HP(s hpq), Red Hat (s rhat) and others are fighting for a piece of the action, although AWS is the market leader. By far.
AWS, thus far, has held to a per-hour pricing model but keeps chipping away at those prices. Amazon CTO Werner Vogels has said the company doesn’t hear demand for by-the-minute pricing. Here’s how Amazon sums up the changes on post to its web site:
- Dedicated Per Region Fee – An 80 percent price reduction from $10 per hour to $2 per hour in any region where at least one Dedicated Instance of any type is running.
- Dedicated On-Demand Instances – A reduction of up to 37 percent in hourly costs. For example, the price of an m1.xlarge Dedicated Instance in the US East (Northern Virginia) region will drop from $0.840 per hour to $0.528 per hour.
- Dedicated Reserved Instances – A reduction of up to 57 percent on the Reserved Instance upfront fee and the hourly instance usage fee. Dedicated Reserved Instances also provide additional savings of up to 65 percent compared to Dedicated On-Demand instances.
Update: As a commenter pointed out, the 80 percent cut is misleading as it is a “per region” fee payable once per hour for each region so does not scale with overall number of instances. “It’s enough to notice but the 37 percent and 57 percent price cuts mentioned are probably more broadly applicable,” another reader said.
Scalr has a good blog post explaining what went on here. The money quote:
“Dedicated Instances are useful if you’re concerned with compliance and security and yet have a use case for the elasticity of the Cloud. They’re a product that’s typically targeted at Enterprises. This move will particularly benefit Enterprises that understand the value of IaaS, and demonstrates yet again that Amazon wants to grab the lion’s share of that expanding market.”
The new pricing will be applied to loads starting July 1, according to the July 9 post. Robert Shear, president of Greystone Solutions, a custom apps shop that uses AWS frequently for customers, said this is a big or not-big whoop depending on your point of view.
“For most customers who are okay with shared physical hardware for pedestrian workloads, it’s not a big deal. Of course, if you’re using dedicated instances, a 37 percent price cut will be welcome,” he said. And, “if you’re a co-lo and you’re quickly running out of arguments for your customers as to why you should continue to exist, I would think that this would turn your fear level from extremely frightened to something even higher.”
Update: This story was updated 7:46 a.m. PDT July 11 with more info and context around the 80 percent price cut and again at 6:57 a.m. PDT Friday July 12 with the Scalr blog link.