Calxeda, Intel And Others Join Facebook's Open Compute Project
Facebook's Open Compute Project is being expanded to incorporate ARM processors, providing new options for companies shopping for low-cost hardware to build out cloud computing environments.
Technologies to bring x86 and ARM processors closer were announced at the Open Compute Summit organized by Facebook in Santa Clara, California. The technologies include the common slot design through which x86 and ARM CPU technologies can be installed on a single motherboard.
Nearly 2,000 people converged on Santa Clara to kick off the fourth OPC Summit on Wednesday. More more than 50 official members and dozens of contributions have joined the Open Compute Project. New organizations include storage companies like EMC, Fusion-io, Hitachi, and Sandisk; microprocessor organizations like Applied Micro, ARM, Calxeda, and Tilera; and communications companies like NTT Data and Orange.
Chip vendors from both the ARM and x86 sides of the house announced they are working together to develop a "common slot architecture" that will allow ARM and x86 processors to coexist side by side on the same motherboard.
"Consumers are smart, we know what we need, we know what we want, we see the current and future challenges," said Frank Frankovsky, Facebook vice president of hardware design and supply chain operations, who opened the event.
The idea is that end customers will be able to design servers that match the workloads they need, at an affordable price, by selecting from the various standardized modules developed by the participants.They would order their servers through systems integrators, who would then work with Taiwanese server vendors to manufacture the systems.
Developing a common slot architecture fits with the goal of reusing as many of the server building blocks developed by the Open Compute Project as possible. The specification will define a processor card, including the CPU, memory and required chipsets, that can then slot into one of the project's motherboard designs.
The common slot design, which Facebook also calls a "group hug board," is being backed by x86 vendors Intel and AMD, as well as ARM vendors AppliedMicro, Calxeda, and Tilera. This specification can be used to produce boards that are completely vendor-neutral and will last through multiple processor generations. The specification uses a simple PCIe x8 connector to link the SOCs to the board.
Calxeda showed Project Knockout on stage. Project Knockout is Calxeda?s ARM-based server motherboard design for the Open Vault storage platform. By installing the Project Knockout board into Open Vault, end users can eliminate the need for a separate server to control the disks in Open Vault.
Calxeda and Avnet are also showcasing work targeting the data center space. The two companies are working jointly to deliver ARM-based contributions to the Open Compute Project that will leverage Avnet?s data center technologies. The co-developed solutions are expected to be available to the OCP community Fall 2013.
At the event, Intel announced a collaboration with Facebook to define the next generation of rack technologies used to power large data centers. As part of the collaboration, the companies also unveiled a mechanical prototype built by Quanta Computer that includes Intel's new photonic rack architecture.
"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. "The disaggregated rack architecture includes Intel's new photonic architecture, based on high-bandwidth, 100Gbps Intel Silicon Photonics Technology, that enables fewer cables, increased bandwidth, farther reach and extreme power efficiency compared to today's copper based interconnects."
Rattner explained that the new architecture is based on more than a decade's worth of research to invent a family of silicon-based photonic devices, including lasers, modulators and detectors using low-cost silicon to fully integrate photonic devices of unprecedented speed and energy efficiency. Silicon photonics is a new approach to using light (photons) to move huge amounts of data at very high speeds with extremely low power over a thin optical fiber rather than using electrical signals over a copper cable. Intel has spent the past two years proving its silicon photonics technology was production-worthy, and has now produced engineering samples.
Silicon photonics made with inexpensive silicon rather than expensive and exotic optical materials provides a distinct cost advantage over older optical technologies in addition to providing greater speed, reliability and scalability benefits.
Intel said it would contribute a design for enabling a photonic receptacle to the Open Compute Project (OCP) and will work with Facebook, Corning, and others over time to standardize the design. The mechanical prototype includes distributed input/output (I/O) using Intel Ethernet switch silicon, and will support the Intel Xeon processor and the next generation, 22 nanometer system-on-chip (SoC) Intel Atom processor, code named "Avoton" available this year.
AMD also launched the AMD Open 3.0 platform (formerly codenamed "Roadrunner.") Powered by the recently announced AMD Opteron 6300 Series processors, AMD Open 3.0, can be installed in all standard 19" rack environments without modification, as well as in Open Rack environments. The AMD Open 3.0 motherboard is a 16" x 16.7" board designed to fit into 1U, 1.5U, 2U or 3U rack height servers. It features two AMD Opteron 6300 Series processors, each with 12 memory sockets (four channels with three DIMMs each,) 6 Serial ATA (SATA) connections per board, one dual channel gigabit Ethernet NIC with integrated management, up to four PCI Express expansion slots, a mezzanine connector for custom module solutions, one serial port and two USB ports. Specific PCI Express card support is dependent on usage case and chassis height.
AMD expects production systems from Quanta Computer and Tyan are expected to be available through Avnet Electronics Marketing, Penguin Computing and other system integrators before the end of Q1.
At the same event, AppliedMicro also announced that it has submitted the first 64-bit ARM-based server specification to the Open Compute Project, based on its upcoming X-Gene server board.
Dell will also back the Open Compute Project systems management framework, although the company is investigating support for the common slot design.
Dell, as well as Hewlett-Packard, are currently offering prototype ARM servers to customers for testing, benchmarking and software development.
At the Open Compute Summit, Dell showed a 64-bit server with AppliedMicro's 64-bit X-Gene ARM chip, which is scalable to 128 CPU cores running at 3GHz. The company already has two 32-bit ARM servers, but is now focusing on 64-bit ARM servers.
"The processor card will be able to support any type of system-on-chip, and it will plug into a backplane that allows you to expand the system out to include a cluster of cards," said Gina Longoria, product marketing director for Calxeda.
The specification still needs to be built out, and no timeframe for it was given Wednesday.
Nearly 2,000 people converged on Santa Clara to kick off the fourth OPC Summit on Wednesday. More more than 50 official members and dozens of contributions have joined the Open Compute Project. New organizations include storage companies like EMC, Fusion-io, Hitachi, and Sandisk; microprocessor organizations like Applied Micro, ARM, Calxeda, and Tilera; and communications companies like NTT Data and Orange.
Chip vendors from both the ARM and x86 sides of the house announced they are working together to develop a "common slot architecture" that will allow ARM and x86 processors to coexist side by side on the same motherboard.
"Consumers are smart, we know what we need, we know what we want, we see the current and future challenges," said Frank Frankovsky, Facebook vice president of hardware design and supply chain operations, who opened the event.
The idea is that end customers will be able to design servers that match the workloads they need, at an affordable price, by selecting from the various standardized modules developed by the participants.They would order their servers through systems integrators, who would then work with Taiwanese server vendors to manufacture the systems.
Developing a common slot architecture fits with the goal of reusing as many of the server building blocks developed by the Open Compute Project as possible. The specification will define a processor card, including the CPU, memory and required chipsets, that can then slot into one of the project's motherboard designs.
The common slot design, which Facebook also calls a "group hug board," is being backed by x86 vendors Intel and AMD, as well as ARM vendors AppliedMicro, Calxeda, and Tilera. This specification can be used to produce boards that are completely vendor-neutral and will last through multiple processor generations. The specification uses a simple PCIe x8 connector to link the SOCs to the board.
Calxeda showed Project Knockout on stage. Project Knockout is Calxeda?s ARM-based server motherboard design for the Open Vault storage platform. By installing the Project Knockout board into Open Vault, end users can eliminate the need for a separate server to control the disks in Open Vault.
Calxeda and Avnet are also showcasing work targeting the data center space. The two companies are working jointly to deliver ARM-based contributions to the Open Compute Project that will leverage Avnet?s data center technologies. The co-developed solutions are expected to be available to the OCP community Fall 2013.
At the event, Intel announced a collaboration with Facebook to define the next generation of rack technologies used to power large data centers. As part of the collaboration, the companies also unveiled a mechanical prototype built by Quanta Computer that includes Intel's new photonic rack architecture.
"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. "The disaggregated rack architecture includes Intel's new photonic architecture, based on high-bandwidth, 100Gbps Intel Silicon Photonics Technology, that enables fewer cables, increased bandwidth, farther reach and extreme power efficiency compared to today's copper based interconnects."
Rattner explained that the new architecture is based on more than a decade's worth of research to invent a family of silicon-based photonic devices, including lasers, modulators and detectors using low-cost silicon to fully integrate photonic devices of unprecedented speed and energy efficiency. Silicon photonics is a new approach to using light (photons) to move huge amounts of data at very high speeds with extremely low power over a thin optical fiber rather than using electrical signals over a copper cable. Intel has spent the past two years proving its silicon photonics technology was production-worthy, and has now produced engineering samples.
Silicon photonics made with inexpensive silicon rather than expensive and exotic optical materials provides a distinct cost advantage over older optical technologies in addition to providing greater speed, reliability and scalability benefits.
Intel said it would contribute a design for enabling a photonic receptacle to the Open Compute Project (OCP) and will work with Facebook, Corning, and others over time to standardize the design. The mechanical prototype includes distributed input/output (I/O) using Intel Ethernet switch silicon, and will support the Intel Xeon processor and the next generation, 22 nanometer system-on-chip (SoC) Intel Atom processor, code named "Avoton" available this year.
AMD also launched the AMD Open 3.0 platform (formerly codenamed "Roadrunner.") Powered by the recently announced AMD Opteron 6300 Series processors, AMD Open 3.0, can be installed in all standard 19" rack environments without modification, as well as in Open Rack environments. The AMD Open 3.0 motherboard is a 16" x 16.7" board designed to fit into 1U, 1.5U, 2U or 3U rack height servers. It features two AMD Opteron 6300 Series processors, each with 12 memory sockets (four channels with three DIMMs each,) 6 Serial ATA (SATA) connections per board, one dual channel gigabit Ethernet NIC with integrated management, up to four PCI Express expansion slots, a mezzanine connector for custom module solutions, one serial port and two USB ports. Specific PCI Express card support is dependent on usage case and chassis height.
AMD expects production systems from Quanta Computer and Tyan are expected to be available through Avnet Electronics Marketing, Penguin Computing and other system integrators before the end of Q1.
At the same event, AppliedMicro also announced that it has submitted the first 64-bit ARM-based server specification to the Open Compute Project, based on its upcoming X-Gene server board.
Dell will also back the Open Compute Project systems management framework, although the company is investigating support for the common slot design.
Dell, as well as Hewlett-Packard, are currently offering prototype ARM servers to customers for testing, benchmarking and software development.
At the Open Compute Summit, Dell showed a 64-bit server with AppliedMicro's 64-bit X-Gene ARM chip, which is scalable to 128 CPU cores running at 3GHz. The company already has two 32-bit ARM servers, but is now focusing on 64-bit ARM servers.
"The processor card will be able to support any type of system-on-chip, and it will plug into a backplane that allows you to expand the system out to include a cluster of cards," said Gina Longoria, product marketing director for Calxeda.
The specification still needs to be built out, and no timeframe for it was given Wednesday.
