Active

While PICMG and its members are committed to the future of ATCA, we recognize that at some time in the future a new, platform that may not be backwards compatible with ATCA may be needed.

A PICMG working group is already defining the requirements for this new platform, which will be called GEN4™.

The GEN4 architecture will be optimized for very high performance computation and network modular platform applications that are beyond the capacities of current AdvancedTCA systems. It will be targeted at high performance computing, central network, network edge, and high capacity storage applications with critical capacity, performance, reliability, density, and efficiency requirements. The explosive growth of Internet traffic (especially driven by video, big data, and the WEB 3.0) will require the deployment of much more capable network elements and will require that they be deployed rapidly. As datacenter and telecommunications networks converge, and cloud computing and various heterogeneous processing models become prevalent, GEN4 will be positioned as the standard modular platform architecture of choice.

GEN4 will be scalable in computational power, networking bandwidth, and storage capacity.

GEN4 will be a new architecture which is complementary to AdvancedTCA. While the boards and shelves will not be plug compatible, the software and management infrastructure elements will be adaptable between architectures. GEN4 will achieve order of magnitude levels of improvement over original PICMG 3.0 R1.0 systems in multiple dimensions, including:

• Architected with energy efficiency, simplicity, and scalability goals set by emerging market demands.
• Module size for high-capacity networking and compute performance using off-the-shelf silicon.
• System throughput (to hundreds of terabits/s), module bandwidth (to tens of terabits/s), and storage capacity in exabytes.
• Efficient power delivery with High Voltage DC options as well as AC
• Module cooling capacity (over 2000 Watts, with fluid cooling options)
• Scalability to efficiently create large multi-frame systems
• Cybersecurity at the lowest level of the hardware architecture to address secure military communications applications.

GEN4 will be a new approach to address the widely varying performance needs of the applications through hierarchy of increasing capability subsystems, supporting SDN and NFV, while providing investment protection for customers with deployed AdvancedTCA systems. And like all PICMG platforms, it will be an open, public standard that anyone can implement.

GEN4 systems will provide the capacity and density required by the next decade of Internet application and traffic growth, serving the modular platform marketplace from 2015 through at least 2025.

This effort is developing enumerated requirements that incorporate 10GBASE-KX4, 10GBASE-KR and 40GBASE-KR4 to the Common Option (ports 0 and 1), Fat Pipes (ports 4-7) and Extended Pipes (8-11) as defined in AMC.2 and used there and in all variants of MicroTCA. A key goal of this activity is to guarantee backward compatibility with existing MTCA and AMC mechanicals and connectors, and interoperability with BX/BX4 fabric options.

In addition to enumerated requirements, the Working Group may generate guidelines/best practices to help ensure uniform interpretation of the specification and ease potential interoperability issues between BX/BX4 and KX/KX4/KR. These best practices may be generated as a separate deliverable, distinct from the enumerated requirements. Deliverables include:

• A 40G Ethernet Specification for MicroTCA.0 and AMC.2 (analogous to PICMG 3.1R2 40G Ethernet for AdvancedTCA)
• Provide a roadmap for higher speed Ethernet Fabrics for the future (100G, etc.).
• Update the MicroTCA guide to include new higher speed Ethernet options.
• Develop test plan and procedure for signal integrity analysis at 40G speeds. Perform tests with shared costs across working group members and PICMG contributing as budget permits.

The AdvancedTCA community is nearing completion of a fairly major enhancement to the core ATCA standard. This new specification is known as “PICMG 3.7” or “ATCA Extensions.” It expands the packaging definitions to include dual sided shelves, where ATCA boards can plug into either the front of the back of a double-deep rack and interconnect through a single or dual backplanes. Both vertical and horizontal board orientations are defined. In addition to this, the Extensions specification also allows for something called Extended Transition Module (ETM) that is essentially a front board-sized circuit board that connects to a front board via Zone 3, much like a standard Rear Transition Module.

These new mechanical configurations significantly expand the flexibility and versatility of ATCA. The double-deep configurations can make better use of available rack space when compared to single-deep systems.

There are many variations of interconnects allowed, but Figure 1 below gives a general idea of the concept.

Importantly, PICMG 3.7 provides a much more detailed definition of, and support for, double wide modules than the original specification. These can support multiple processors, bigger heatsinks, cheaper full height memory modules, and multiple disk drives on a single assembly if desired. PICMG 3.7 also defines requirements for typical data center environments in addition to the telecom central office. Double wide modules can support up to 800W of power dissipation if the shelf is built for that. AC as well as traditional -48VDC power environments are also supported. PICMG 3.7 also provides a much more rigorous definition of the thermal environment within a shelf and provides useful guidance for those designing high power systems.