Decoupling refresh cycles lets data center customers purchase server compute nodes, which are essentially very small motherboards (processor, system memory, boot environment, management, and connection to a local network fabric), separately from rack-level data center network interfaces (1GbE and 10GbE), separately from mass storage (hard disks and solid-state drives).

FORBES: AMD SeaMicro Disaggregation: A Threat to Facebook's OCP?

The AMD Open 3.0 platform is designed to easily enable IT professionals to "right size" the server to meet specific compute requirements.

ENGADGET: AMD unveils Open 3.0: an Opteron 6300 platform for the Open Compute Project

Businesses with large data centers can significantly reduce capital expenditure by disaggregating or separating compute and storage resources in a server rack.

ENGADGET: Facebook's Open Compute Project splits up monolithic servers with help from Intel, more

The Silk browser software resides both on Kindle Fire and on the massive server fleet that comprises the Amazon Elastic Compute Cloud (Amazon EC2).

ENGADGET: Amazon Silk browser spins a faster mobile web, courtesy of cloud servers (video)

Ordinary databases and grid may be too slow, but ScaleOut Software stores data in memory on server farms or uses distributed caching on compute grids, to eliminate bottlenecks and reduce latency.

FORBES: Big Data Analyzed Fast In Memory -- For Finance, Online Dating, Home Shopping Network

"Intel and Facebook are collaborating on a new disaggregated, rack-scale server architecture that enables independent upgrading of compute, network and storage subsystems that will define the future of mega-datacenter designs for the next decade, " said Justin Rattner, Intel's chief technology officer during his keynote address at Open Computer Summit in Santa Clara, Calif.

ENGADGET: Facebook's Open Compute Project splits up monolithic servers with help from Intel, more

"Cloud computing has brought a sea change to the data center--dramatically altering the economics of compute by changing the workload and optimal characteristics of a server, " said Andrew Feldman, SeaMicro CEO, who will become general manager of AMD's newly created Data Center Server Solutions business.

ENGADGET: AMD absorbs server startup SeaMicro for $330 million, says it's no impulse buy

Zuckerberg takes stage: Today we want to talk about infrastructure, server design, and data design behind Facebook called Open Compute Project.

FORBES: Live From Facebook's "Open Compute Project" Announcment

The FirePro S10000 is merely the latest salvo: a dual-GPU, server-focused version of the W9000 that greatly increases overall compute power, delivering 5.91 TFLOPs of single precision calculations and 1.48 TFLOPS of dual precision performance in a single PCIe 3.0 card with 6GB of GDDR5 RAM.

ENGADGET: AMD's dual-GPU FirePro S10000 gobbles watts, spews out nearly 6 TFLOPs for server graphics

Server optimization features such as SSL termination and compression technologies use specialized hardware to offload compute-intensive operations from the servers to the application delivery controllers.

FORBES: For IT Managers, Going Green Can Save You Some Long Green