At re:Invent last year, Amazon VP James Hamilton gave attendees a rare glimpse into the inner workings of the AWS juggernaut, revealing how his firm leverages hyperspecialized components to achieve operational efficiencies at unprecedented scale. But other than that, the IaaS giant has remained tight lipped about its cloud, not disclosing revenues let alone providing any concrete details about the technologies that power its services.

Therefore, it comes as no surprise that rival Microsoft is doing the exact opposite and making its server and rack blueprints freely available through the Open Compute Project, which was established by Facebook in 2011 to accelerate the adoption of hyperscale architectures. Bill Laing, the head of the software giant’s enterprise cloud business, detailed in a Monday blog that Redmond is looking to leverage its technical know how in hopes of gaining a bigger presence in the open source community and ultimately the data center.

According to Laing, Microsoft’s server costs 40 percent less than standard systems, runs 15 percent more efficiently and can be deployed twice as fast with improved service times. Additionally, it enabled the company to reduce network cabling by 1,100 miles and metal by 10,000 tons across its base of one million machines.

The design was released along with the platform Microsoft uses for infrastructure operations such as diagnostics and power and cooling. As Windows Server and System Center VP Brad Anderson explained in a recent interview, the technology was born entirely out of necessity.

“In the last several years we’ve spent more than $15 billion as we built out Azure, the data centers, all of the things that are inside the data centers and hosting these 200-plus services like Bing, like Skype, like Outlook.com,” Anderson detailed. “When you’re spending $15 billion over a couple of years, you’re bringing in hundreds of thousands of servers at a time, making 50,000 networking changes a day; you have to innovate like crazy.”