The X Factor
Camera Link HS: The Path to 50Gbps and Beyond
Inexpensive IP core speeds development and implementation of Camera Link HS for camera and frame grabber manufacturers. CLHS is poised to provide significantly expanded capability (50Gbps and beyond) in upcoming releases.
Bild: Alysium Tech

Camera Link High Speed (Camera Link HS or CLHS) is a robust interface standard for industrial cameras and systems. CLHS was introduced in 2012 and built on the low-latency, zero-jitter, real-time strengths of Camera Link. The CLHS standard further provided scalable bandwidths with reliable data delivery, support for both copper and fiber cabling, and, importantly, Intellectual Property (IP) cores for ease of component development. Currently at version 1.1, CLHS implements its protocol with two physical layer encodings: M Protocol and X Protocol. The VHSIC Hardware Description Language (VHDL) IP cores available from A3 have been implemented in Altera, Xilinx, and Microsemi FPGAs. Used in components such as cameras and frame grabbers, the cores are available from A3 at a price point of $1,000 to help promote the standard and speed proliferation of the technology. The inexpensive IP core infrastructure helps camera and frame grabber manufacturers speed CLHS development and implementation. Because CLHS IP core code is open-source, licensed users can customize the core for their own use. They can improve it by adding features that, upon technical review and approval, can become a permanent part of the IP core.

The CLHS M Protocol IP core features 8b/10b line encoding for use in lower-cost and lower-power FPGAs. It supports hardware with copper (C2) cabling, with CX4 connections up to 15m and 2.1Gbps in the CLHS cable. Active plug-on fiber cable can be added externally to achieve up to 100m. Typical bit rates for the M Protocol are up to 5Gbps per lane. The CLHS X Protocol IP core is designed for 10Gbps and higher bit rates using 64b/66b line encoding. Other interface protocols, such as CoaXPress (CXP), are moving to this encoding structure. Note that because of the line encoding technique, CLHS 10Gbps is equivalent to CXP 12.5Gbps. The X Protocol also has forward error correction and single-bit error immunity. This protocol is designed to natively support fiber connections F2 (SFP+) and soon to be introduced F3 – quad small-form-factor pluggable (QSFP+) and multifiber push-on (MPO) – connectors at rates up to 16Gbps per lane. Copper CX4 (C3) connections still are supported, but the highest data rates and cable lengths (as much as 20km or more, in some cases) are realized using fiber cabling. Importantly, the 64b/66b encoded X Protocol has been in production since 2012 and has a long history of field-tested stability.

Image 2 | The typical X-Protocol (XP) Camera Link HS device architecture features the XP core at the front end, which bridges user logic to the CLHS PCS layer IP.
The typical X-Protocol (XP) Camera Link HS device architecture features the XP core at the front end, which bridges user logic to the CLHS PCS layer IP. – Bild: Association for Advancing Automation (A3)

CLHS M/X Protocol IP cores

Five message types are available in CLHS: a low-latency (<1µs) 0ns jitter trigger for real-time imaging at up to 8MHz with seven separate modes, 16 bidirectional general-purpose I/Os with <1µs latency, remote DMA video packet and frame grabber data forwarding, high-speed command uplink enabling streaming correction coefficients, and revision information used in device discovery. One of the big values of CLHS is that it is a real-time, point-to-point interface. In the context of the CLHS triggers and GPIO, the low latency means immediate responses, eliminating the need for tracking and scheduling as might be the case with other interfaces. The CLHS X Protocol can connect to a 10Gbase-KR Ethernet PHY physical (Figure 2) for FPGAs that don’t have built in 10Gbps transceivers or use the provided Physical Coding Sublayer (PCS) module to implement the code within 10Gbps-capable FPGAs that don’t offer the 10Gbase-KR capability (Figure 3). Because of the unique way the CLHS X Protocol is implemented, Camera Link HS is already in the final stages of moving the specification to 25Gbps.

Ready for 25Gbps

The well-designed IP core architecture makes CLHS ready without any modification to move to 25Gbps on an FPGA using 25Gbps transceivers over fiber with SFP28 or QSFP28 or MPO connectors. The soon-to-be-released specification revision will validate CLHS as having up to 100Gbps (4×25Gbps) available bandwidth using a single MPO connector. In achieving this speed increase, the CLHS X Protocol IP core remains completely unchanged. Therefore, all existing CLHS infrastructure also remains unchanged, making the move to 25Gbps inexpensive and straightforward for vendors and users. Contrast this with frame grabber interfaces that need to change core code and move to different line encoding to implement faster rates. Furthermore, CLHS has a very flexible and robust mechanism for changing the speed of the interface. Using MPO to LC, breakout cables, or SFP28 allows backward compatibility to existing CLHS 10Gbps frame grabbers or newly developed 25Gbps frame grabbers to support existing 10G cameras.

The CLHS X Protocol can use the provided physical coding sublayer (PCS) module to implement code within 10Gbps frame grabbers.
The CLHS X Protocol can use the provided physical coding sublayer (PCS) module to implement code within 10Gbps frame grabbers. Bild: Association for Advancing Automation (A3)

Getting to 50Gbps and Beyond

With FPGA and PHY technology ready to support 50Gbps and higher rates, CLHS is poised to deliver powerful real-time camera interfacing at even higher imaging rates. The features and capabilities already in place in the CLHS standard stand to make the path to 50Gbps effortless. The CLHS working group has leveraged the existing X Protocol IP core, adding a simple bridge from the 64bit output of the X Protocol IP core PCS to the 128bit input of 50Gbps FPGAs with built-in SerDes blocks. To achieve 50Gbps speeds, CLHS will introduce a virtual channel using a master and slave channel (or lane) all in one fiber. With this minor change in the specification, component vendors will be able to easily migrate to the higher-speed interfacing with minimal impact on hardware and software architecture.

Future of Frame Grabbers

The Camera Link HS standard is built for the future of imaging in its implementation of general IP cores, open-source VHDL, and inherent advanced line encoding and error correction. One key to future-proofing the standard is the seamless support for fiber cabling, something that may become standard in most imaging environments. Fiber cable is not fragile or more susceptible to bending, wrapping, or pinching than traditional copper wire. With current technologies, fiber is easy to work with and is field installable and modifiable. Able to carry more than twice the bandwidth of copper at distances of many kilometers, fiber is also less expensive than copper wire. Already a robust imaging interface, Camera Link HS continues to improve. Its use of 64b/66b encoding in 2012 still provides manufacturers and end users a long design cycle from 10Gbps all the way up to 50Gbps. The standard’s many features and capabilities make it stand out as a definitive choice in high-speed imaging applications both now and in the future.

www.automate.org
Association for Advancing Automation (A3)
http://www.automate.org

Das könnte Sie auch Interessieren

Bild: Kontron Europe GmbH
Bild: Kontron Europe GmbH
Democratizing AI

Democratizing AI

Der Einstieg in die AI fällt vielen Unternehmen noch schwer, nicht zuletzt wegen der komplexen Hardware-Auswahl. In einer AI-Workstation verbinden Robotron und Kontron nun zwei Lösungsansätze zur unkomplizierten Anwendung von Computer Vision. Die Workstation ermöglicht auch Einsteigern das Trainieren, Ausführen und Inferenz der Algorithmen für eine Vielzahl von Modellen und Frameworks.

Bild: IDS Imaging Development Systems GmbH
Bild: IDS Imaging Development Systems GmbH
Verpixelte Gesichter

Verpixelte Gesichter

Dort wo personenbezogene oder andere sensible Daten bei Computer Vision oder kamerabasierter Bildverarbeitung im Fokus stehen, spielt das Thema Datenschutz eine große Rolle. Abhilfe soll eine dezentrale Datenverarbeitung durch Edge Computing schaffen. Ohne die Privatsphäre einzelner zu verletzen, sollen prozessrelevante Informationen direkt im Gerät extrahiert oder sensible Bildbereiche unkenntlich gemacht werden, bevor Aufnahmen das Gerät verlassen und weiter verarbeitet werden.

Bild: Vision Components GmbH
Bild: Vision Components GmbH
OEM Stereo-Kamera

OEM Stereo-Kamera

Wie Entwickler ihre Produkte schneller an den Markt bringen, Kosten einsparen und Designrisiken minimieren können, hat Vision Components kürzlich mit dem VC Power SoM
gezeigt, einem FPGA-basierten Hardwarebeschleuniger für Embedded Vision.
Jetzt entwickelt das Unternehmen sein FPGA-Modul weiter und präsentiert die VC Stereo Cam für 3D- und Zweikamera-Anwendungen.

Anzeige

Anzeige

Anzeige

Anzeige

Anzeige