Nuvation HEADLINES 
|
New IP
|
| » |
ATA 4/5, UDMA 33/66 Core for Xilinx Spartan 3 |
| » |
ATA 4/5, UDMA 33/66 for Altera Cyclone |
| » |
GFP-F IP Core |
| » |
PCI-Express Core |
|
|
New Affiliations
|
| » |
W&W Communications H.264 CODEC Integration Partner |
| » |
Lattice Certified FPGA Design Center |
| » |
Intel PCI Express Developers Network |
| » |
ADI Certified DSP Partner |
|
|
Win a
Free Moose!
We're giving away this cute, plush, beanie Nuvation Moose for your feedback and comments on this newsletter. Tell us!
|
|  |
|
|
ATCA - The NexGen Telecom Standard
Advanced Telecom Computing Architecture
 What is ATCA?
Background
ATCA Feature Overview
CompactPCI vs. AdvancedTCA
ATCA Mechanical Configuration
ATCA Power & Cooling
ATCA Shelf Management
ATCA Data Transport
Designing an ATCA Board
ATCA in your Product: Free Consultation
|
|
What is ATCA?
|
ATCA stands for Advanced Telecommunications Computing Architecture, and is a new open industry specification for designing high-performance telecom and datacom applications. It has been developed by the PICMG group with the goal of providing "an open platform standard that meets the needs of telecom infrastructure equipment for the next ten years." (Source: PICMG) The ATCA platform supports multiple switch fabrics, with enhanced mechanics, backplane, system management, power distribution, and cooling.
|
|
Background
|
  The telecommunications industry is pushing the limits of existing standards. New silicon runs faster and packs in more features, but requires more board area and spacing, more power and cooling. The backplane capacity is limited, and the chassis mechanical standards don't match telecom equipment frames and required I/O. There are bus limitations and power distribution issues with existing standards like VME and CompactPCI. At the same time, network data rates are increasing to the point where performance is limited by I/O issues. Voice and data convergence is being implemented, enhanced service platforms are increasingly utilizing standard computers, and telecom giants are outsourcing more of their electronic equipment design.
From this situation, ATCA emerges as the standard that meets all the criteria, and is designed not as a general computer platform like PCI and the likes, but specifically for telecommunications equipment. Evolving since 2001, the ATCA platform has been developed by PICMG (www.picmg.org) and over 100 participating companies to form the PICMG 3.0 specification (www.advancedtca.org).
|
|
ATCA Feature Overview
|
- Serial links & switch fabric technology, eliminating failures to multiple spots from a single failure in a parallel bus scheme
- Large board (~11"x13") to support new silicon with more power and up to 200W of cooling per board
- Chassis based on 600mm (~24") ETSI frame equipment practice, with provisions of 19" and 23" rack versions
- 2 to 16 boards per chassis
- Hot-swap capability on all boards and active modules
- Dual-redundant Ethernet control plane (independent of data transport switch fabric)
- I/O features tailored for Telecom specific applications:
- Multi-protocol support for interfaces up to 40Gb/s
- Front-to-back clearances that allow simultaneous front and rear I/O taking to account connector bodies and fiber bend radii
- Optional rear transition module that supports rear I/O
- Single board failure domain (no parallel buses)
- Dual redundant -48 VDC power feeds
- Sophisticated system management, with electronic keying, module power control, health monitoring, and active cooling control
- Up to 2.4 Tb/s aggregate bandwidth per chassis when full mesh topology is used
- Support for industry-standard mezzanine modules
|
|
CompactPCI vs. AdvancedTCA
|
|
Attribute
|
CompactPCI (PICMG 2)
|
AdvancedTCA (PICMG 3)
|
|
Board Size
|
6U x 160mm x 0.8"57 sq in + 2 Mez
|
8U x 280mm x 1.2"140 sq in + 4 Mez
|
|
Board Power
|
35-50W
|
150-200W
|
|
Backplane Bandwidth
|
~4Gb/s
|
~2.4Tb/s
|
|
#Active Boards
|
21
|
16
|
|
Power System
|
Central converter5, 12, 3.3 V Backplane
|
Distributed converterDual 48V Backplane
|
|
Management
|
OK
|
Advanced
|
|
I/O
|
Limited
|
Extensive
|
|
Clock, update, test bus
|
No
|
Yes
|
|
Regulatory conformance
|
Vendor Specific
|
In Standard
|
|
Multi-Vendor support
|
Extensive
|
Growing
|
|
Base cost of shelf
|
Low
|
Moderate
|
|
Functional density of shelf
|
Low
|
High
|
|
Lifecycle cost per function
|
High
|
Low
|
Source: Chuck Byers, Lucent Technologies
AdvancedTCA does not use parallel buses (like the PCI Bus), used by CompactPCI or VME. A single failure in a parallel bus can affect multiple slots. Power distribution is simplified by the use of redundant power. Additionally, the board area and spacing is larger in the ATCA standard, allowing for larger heatsinks needed for cooling down today's CPU's.
|
|
ATCA Mechanical Configuration
|
- 8U boards in 12U chassis
- 1.2" board pitch allows heat sinks plus rear SMT
- Forced air cooling for up to 200 watts per slot
- Front and rear fiber bend area in 600mm depth
- Simplified sheet metal construction
- ETSI & NEBS vibration, shock and serviceability
|
|
ATCA Power & Cooling
|
- -48V/-60 VDC power input
- Redundant power inputs
- Distribution of ringing voltages
- Capacity of over 3,200 Watts per shelf
- Local power conversion on each board
|
|
ATCA Shelf Management
|
- Monitor & control low-level aspects of ATCA boards and other field replaceable units (FRUs) within a shelf
- Watch over basic health of the shelf, report anomalies, take corrective action when needed
- Retrieve inventory information & sensor readings
- Receive event reports and failure notifications from boards and other intelligent FRUs
- Manage power, cooling & interconnect resources in the shelf
- Enable visibility into a shelf for a logical System Manager
- Overall, sophisticated shelf management
|
|
ATCA Data Transport
|
- ATCA backplane is designed for four different fabric topologies including Star, Dual Star, Dual-Dual Star, and Full Mesh
- In the Star topology, each node board is connected to one central switching board. Applications for the star topology include non-carrier-grade with little latency sensitive data traffic.
- In the Dual-Star topology, there are 2 connected redundant switching boards that are centralized, and each node board connects to both switching boards. Redundancy decreases downtime from failure. Applications for the dual-star topology include carrier-grade with non-latency-sensitive data requirements like a modular server.
- In the Dual-Dual Star topology, there are 2 distinct and redundant (i.e 4 in total) centralized fabrics, one for control and one for data. Each node board is linked to all four of the fabrics. This topology is useful for carrier grade applications with latency sensitive streaming data requirements and significant control and management, TCP/IP-based, workload. Data throughput is optimized since the control and data planes are separated.
- In the Full Mesh topology, each board can be directly connected to every other board. Switching and management is distributed across all boards, with no centralized switching board. This topology is suited for carrier-grade applications with large data throughput requirements, like routers. The full mesh topology is highly redundant and scalable, as all boards can intercommunicate simultaneously. A 16-slot ATCA shelf with a full mesh topology of 10Gbps per connection has a total bandwidth of 2.4 Tbps!
- Fabric alternatives defined for ATCA (within PICMG 3.x specifications) include Ethernet, FiberChannel, InfiniBand, StarFabric, PCI-Express, and Rapid I/O
|
|
Designing an ATCA Board
|
Depending on the fabric topology being implemented, there are a variety of boards that Nuvation can custom design for your application:
Front Boards:
- Mesh-enabled boards for the full-mesh topology with connections on each board to 2 to 15 other boards in the chassis
- Node boards for star and dual star topologies
- Hub boards to provide centralized switching for star and dual star topologies
- Base interface boards for applications that only require the base Ethernet interconnectivity and no fabric interface for switching. Note: All other front boards have the base Ethernet interface as well
Rear Transition Modules (RTMs):
RTMs are optional boards that can be designed to connect directly to any front board with custom interfaces and connectors between the two boards. The general purpose is to implement all I/O cable assemblies on the RTM, and all active components on the front board, allowing for easier servicing on the front board without the need for disconnecting and reconnecting cables. The ATCA specification defines space for the RTM in the chassis and is open ended on whether these boards are PCBs, wiring harnesses, or other constructions.
Mezzanine Cards:
ATCA front boards can connect up to 4 current PMC (PCI Mezzanine Card) modules. However, with the emerging need for speed, space and cooling, the form factor of the PMC modules has room to increase due to the larger size of the ATCA front boards.
ATCA Backplanes:
The ATCA backplane has 3 zones. As seen in the illustration Zone 1 is for power and system management.
Zone 2 is for the base interface and switch fabric. Zone 3 is for front board to RTM connection.
|
ATCA in your Product: Free Consultation
|
ATCA is an ideal platform for NexGen multi-service switches that require high-capacity backplanes that support control traffic plus packet and cell data traffic with well-defined quality of service (QoS). Nuvation can design your ATCA front boards, rear boards, and mezzanine cards for next-generation DSLAMs, carrier-grade media servers, and non-carrier grade applications. If you are in the process of upgrading your existing equipment to ATCA standards or need help with ATCA in your current product designs, contact sales@nuvation.com for a free consultation.
ATCA Specification details courtesy of PICMG
|
|
Customer service
· To subscribe yourself or a friend, please click
here.
· Questions? Comments?
Send us your feedback.
|
|
 |
