background preloader

Open Compute Project

Facebook Twitter

Hacking Conventional Computing Infrastructure
We started a project at Facebook over two years ago with a pretty big goal: to build one of the most efficient computing infrastructures at the lowest possible cost.

We decided to honor our hacker roots and challenge convention by custom designing and building our software, servers and data centers from the ground up – and then share these technologies as they evolve.

The result is a data center full of vanity free servers which is 38% more efficient and 24% less expensive to build and run than other state-of-the-art data centers.

Why Open Hardware?
By releasing Open Compute Project technologies as open hardware, our goal is to develop servers and data centers following the model traditionally associated with open source software projects. That’s where you come in. Open Rack » Open Compute Project. Project Chair: Matt Corddry The Open Rack is the first rack standard that’s designed for data centers, integrating the rack into the data center infrastructure, part of the Open Compute Project’s “grid to gates” philosophy, a holistic design process that considers the interdependence of everything from the power grid to the gates in the chips on each motherboard.

You can download the mechanical drawings for the v1.0 single column rack, the v0.6 triplet rack and the v1.0 PDU. You can also download the latest Open Rack specification and the latest Open Rack design guide. Optical Interconnect for Open Rack Intel has submitted a design guide for OCP Open Rack adopters that provides an overview for implementing an intra-rack optical interconnect scheme that utilizes a New Photonic Connector (NPC) and embedded optical modules to deliver next generation data rate scaling and system architectural benefits.

You can download the design guide. Deploying OCP in a Collocated Facility. Solution Providers » Open Compute Project. When we started the Open Compute Project and shared our design specifications with the community, we envisioned that you would take the designs and make efficient, scale out solutions. Many of you have. Another way to benefit from the Open Compute Project is to work with an Open Compute Solution Provider to design and purchase a system. A Solution Provider (SP) can work with you to determine the right mix of OCP technologies to meet your project needs. They can help you fully realize the efficiencies of an OCP solution by bringing together the synergies of the data center, hardware platform and application. When dealing with an SP, you can expect a high level of technical expertise, an efficient supply chain, and capable support services as they deliver rack level solutions for your data center deployments.

Feel free to contact our Open Compute Project Solution Providers so you can bring the efficiencies of the Open Compute Project to you data center! Networking » Open Compute Project. Project Chairs: Omar Baldonado | Carlos Cardenas The Open Compute Networking Project is creating a set of technologies that are disaggregated and fully open, allowing for rapid innovation in the network space. We aim to facilitate the development of network hardware and software – together with trusted project validation and testing – in a truly open and collaborative community environment.

We’re bringing to networking the guiding principles that OCP has brought to servers & storage, so that we can give end users the ability to forgo traditional closed and proprietary network switches - in favor of a fully open network technology stack. Our initial goal is to develop a top-of-rack (leaf) switch, while future plans target spine switches and other hardware and software solutions in the space.

Hardware Management » Open Compute Project. Compliance and Interoperability » Open Compute Project. Project Chair: YF Juan As the Open Compute Project continues to grow, and as the community submits more and more content, there is a great need for a transparent and scalable approach that ensures solutions meet minimum OCP standards for compliance and interoperability and are branded as such. The OCP Compliance & Interoperability project (C&I) will establish a framework that enables a transparent approach for qualifying solutions to become OCP branded.

It will allow providers to leverage a common set of tools and processes to deliver solutions that meet OCP defined guidelines and expectations and consumers the assurance that OCP branded solutions meet those expectations. In addition it positions OCP C&I as a base certification layer that can be leveraged by communities & technologies beyond OCP. The charter is available for download.

Data Center » Open Compute Project. Storage » Open Compute Project. Project Chair: Asghar Riahi and Per Brashers Storage is a key component of any data center, and offers many opportunities for efficiency gains. Here are some contributions to the storage project: Cold Storage As the requirements for cold data — where data is stored on disk but almost never read again, like legal data or backups of third copies of data — keep on increasing dramatically, there are huge demands for developing some form of cold storage system with the highest capacity and the lowest cost.

Cold storage is designed as a bulk load fast archive. You can download the cold storage specification. Fusion-io The Fusion-io 3.2 TB I/0 is a high-density I/O PCI Express adapter card. You can download the specification and CAD models for the 3.2TB I/O card. Hyve The Torpedo design concept is a 2xOpenU storage server that can accommodate 15 3.5″ drives arranged in a 3 x 5 array. The initial design target is direct cabled to an off the shelf SAS expander. OpenNVM Open Vault. Server Design » Open Compute Project. Project Chair: Mark Shaw Open Compute motherboards are power-optimized, barebones designs that provide the lowest capital and operating costs. Many features found in traditional motherboards have been removed from the design. The OCP specification for the AMD v2.0 motherboard is a dual AMD G34 socket motherboard (Magny Cours or Interlagos) with 16 DIMM slots (PCB mechanical and motherboard mechanical coming soon). AMD developed their own version of the board, the AMD Open 3.0 modular server, for the AMD Financial Services Open Platform.

The OCP Asia-Pacific chapter translated the AMD v2.0 motherboard into Japanese. Calxeda has developed an ARM-based server that converts the Open Vault JBOD into a storage server. High Availability The high availability (HA) server leverages the Intel Motherboard Hardware Specification v2.0. Intel The OCP specification for the Intel v2.0 motherboard uses two Intel® Xeon® E5-2600 processors and 16 DIMM slots per board. System on a Chip Power Supply. Home » Open Compute Project.