High speed OC-768 configurable link layer chip

The present invention relates to network chip architecture generally and, more particularly, to a high speed OC-768 configurable link layer chip.

Each channel of a conventional customer configurable network interface utilizes large field programmable gate arrays (FPGAs) for the media access controller (MAC) layer and glue, application specific integrated circuits (ASICs) or FPGAs for the forward error correction (FEC) layer and standard products for the serializer/de-serializer (SERDES). The conventional architecture enables a fairly rapid time to market for custom solutions, but at a high cost, a large board area and a large power budget.

OC-768 devices will be fielded in wide area network (WAN), storage area network (SAN), and local area network (LAN) applications. A fast time to market is desirable in all the applications. However, each application can have slightly different configurations. For example, in the WAN application, each optical customer can utilize a proprietary forward error correction (FEC) algorithm, while in the SAN application different manners of packet filtering can be desirable on each channel.

It would be desirable to provide a platform that is rapidly customizable for the portions of logic that vary between the WAN, SAN, and LAN applications.

The present invention concerns an integrated circuit comprising a plurality of link layer controllers. The plurality of link layer controllers may be configured to operate independently in a first mode and cooperatively in a second mode.

The objects, features and advantages of the present invention include providing a high speed OC-768 configurable link layer chip that may (i) have a reasonable power budget, (ii) be implemented in a single package, (iii) provide an integrated design, (iv) provide a single base architecture that satisfies multiple markets, (v) provide a single base architecture that satisfies multiple slightly different applications of a single customer, (vi) reduce engineering costs, (vii) provide easier parts management and/or (viii) provide a solution where the customer designs only the customer value added areas.

These and other objects, features and advantages of the present invention will be apparent from the following detailed description and the appended claims and drawings in which:

FIGS. 1A and 1B are block diagrams of example configurations of a preferred embodiment of the present invention;

FIG. 2 is a block diagram of an example application in accordance with a preferred embodiment of the present invention;

FIG. 3 is a block diagram of a preferred embodiment of the present invention;

FIG. 4 is a detailed block diagram of another preferred embodiment of the present invention;

FIG. 5 is a block diagram of yet another preferred embodiment of the present invention;

FIG. 6 is a more detailed block diagram of an example application of the embodiment of FIG. 5 ;

FIG. 7 is a more detailed block of an example implementation of the embodiment of FIG. 5 ;

FIG. 8 is a block diagram of various example channel implementation in accordance with the present invention;

FIG. 9 is a block diagram of an implementation of the embodiment of FIG. 1A ; and

FIG. 10 is an alternative embodiment of the embodiment of FIG. 1A .

Referring to FIGS. 1A and 1B , block diagrams of a circuit 100 are shown in accordance with a preferred embodiment of the present invention. The circuit 100 may be implemented, in one example, as a link layer chip. The circuit 100 may be configured, in a first mode, as a 40 Gb/s device ( FIG. 1A ) that may run as, in one example, an OC-768 compliant link layer chip and, in a second mode, as a quad 10 Gb/s device ( FIG. 1B ). The circuit 100 may be rapidly customized to support different applications in storage area networks (SAN), wide area networks (WAN) and/or large area networks (LAN). The circuit 100 may be configured to support a number of communication protocols (e.g., WIS, SPI-5, SPI-4, XGMII, XAUI, etc.). The circuit 100 may provide user customizable logic areas that may be configured to provide a number of custom (or proprietary) functions (e.g., forward error correcting (FEC), packet filtering, host interface bus protocol management, etc.). The circuit 100 may be implemented as a single integrated circuit. The circuit 100 is generally configured to couple a host device (or system) to a synchronous optical network (SONET). In one example, the circuit 100 may be configured to provide an OC-768 compliant interface.

Referring to FIG. 2 , an example application in accordance with a preferred embodiment of the present invention is shown. In one example, a number of circuits 100 may be coupled together to provide a link in a high speed optical network (e.g., OC-768) via a number of lower speed links (e.g., 10 Gb/s Ethernet). The circuits 100 may be configured to distribute data from the high speed network on the number of lower speed links and recombine the data from the lower speed links for presentation to the high speed network.

Referring to FIG. 3 , a detailed block diagram of a circuit 100 in accordance with a preferred embodiment of the present invention is shown. In one example, the circuit 100 may comprise a circuit 102 , a number of circuits 104 a–n and a number of circuits 106 a–n . The circuit 102 may be implemented as an interface controller. The circuits 104 a–n may be implemented as link layer controllers. The circuits 106 a–n may be implemented as serializer/deserializer (SERDES) circuits.

The circuit 102 may be configured to implement one or more communication protocols (e.g., SPI-4, SPI-5, etc.). The circuit 102 may have an input 110 that may receive one or more control signals (e.g., CTRL — IN), an input 112 that may receive one or more data signals (e.g., DATA — IN), an output 114 that may present one or more data signals (e.g., DATA — OUT) and an output 116 that may present one or more control signals (e.g., CTRL — OUT). The interface controller 102 is generally coupled to the circuits 104 a–n . The circuit 102 may be configured to operate the circuits 104 a–n cooperatively as a single high speed communication channel, in a first mode, and independently as a plurality of lower-speed communication channels, in a second mode.

The circuits 104 a–n may be implemented, in one example, as media access controllers (MACs). Alternatively, the circuits 104 a – 104 n may be implemented as framers. However, other types and/or combinations of link layer controllers may be implemented accordingly to meet the design criteria of a particular application. Each of the circuits 104 a–n is generally coupled between the circuit 102 and a respective one of the circuits 106 a–n . The couplings (or interfaces) between the circuits 104 a–n and the circuits 102 and 106 a–n are generally customizable (or configurable) to meet the design criteria of a particular application.

The circuits 106 a–n may be implemented, in one example, as serializer/deserializer (SERDES) circuits. Each of the circuits 106 a–n generally has a respective input/output interface 118 a–n . The circuits 106 a–n may be configured to support one or more communication protocols (e.g., WIS, SPI-5, SPI-4, XGMII, XAUI, etc.).

Referring to FIG. 4 , a detailed block diagram of a circuit 100 ′ is shown illustrating an alternative embodiment of the present invention. The circuit 100 ′ may comprise an interface controller 102 ′, a number of link layer controllers 104 a ′– 104 n ′, and a number of SERDES 106 a ′– 106 n . The circuit 100 ′ may be implemented similarly to the circuit 100 except that each of the link layer controllers 104 a ′– 104 n ′ may be coupled to the respective serializer/deserializer 106 a ′– 106 n ′ via a respective customizable logic block 120 a–n . The customizable logic blocks 120 a–n may be configured, in one example, to provide proprietary (user specific) interface circuitry (e.g., forward error correction (FEC), packet filtering, etc.). The customizable logic blocks 120 a–n may be implemented with one or more diffusion layers that may comprise structures (e.g., A-cells, standard cells, hard macros, sea-of-gates, gate array, etc.) that may form transistors, logic, memory and/or other custom circuitry with the addition of one or more metal layers (e.g., wiring, routing, etc.). In one example, the customizing metal layers may be deposited at a different time (e.g., subsequent manufacturing step) than the fabrication of the circuit 100 ′. For example, the circuit 100 ′ may be fabricated and stored for customization at a later date.

Referring to FIG. 5 , a block diagram of a circuit 100 ″ is shown illustrating yet another preferred embodiment of the present invention. The circuit 100 ″ may comprise a circuit 102 ″, a number of circuits 104 a ″– 104 n ″, a number of circuits 106 a ″– 106 n ″, a number of customizable logic blocks 120 a ′– 120 n ′ and a user customizable logic block 122 . The circuit 100 ″ be implemented similarly to the circuit 100 ′ of FIG. 4 except that the user customizable logic block 122 may couple the circuit 102 ″ and the circuits 104 a ″– 104 n ″. The customizable logic blocks 120 a ′– 120 n ′ and 122 may be implemented similarly to the blocks 120 a–n (described above in connection with FIG. 3 ). For example, the user customizable logic blocks 120 a ′– 120 n ′ and 122 may comprise gates, memory, etc. that may allow for flexible designs such as statistic gathering for the link layer controllers 104 a ″– 104 n ″. In general, the customizable logic blocks 120 a ′– 120 n ′ and 122 provide for user definable customization of the interactions between the circuits 102 ″, 104 ″– 104 n ″ and 106 a ″– 106 n ″ to meet the design criteria of particular applications.

The customizable block 122 may provide a number of user definable input/outputs 124 . In general, the present invention may define a set of input/outputs, memory and logic that provides standard options of a link layer chip design while leaving room for customization to include user specific elements, features and/or functions. For example, customized (or proprietary) error detection, tracking and/or correction may be implemented via the customizable logic blocks 120 a ′– 120 n ′ and 122 .

Referring to FIG. 6 , a more detailed block diagram of an example implementation of the circuit 100 ″ of FIG. 5 is shown. In one example, the circuit 102 ″ may comprise a control pin logic and I/O block 130 , a number of banks of quad serializer/deserializers 132 , a configurable multiplexer circuit 134 that may be configured to select between a number of protocols (e.g., SPI-5, SPI-4, link layers using SPI-5 narrow mode, etc.), an SPI-5 link layer controller 136 and a number of (e.g., four) SPI-4 link layer controllers 138 . The circuit 100 ″ may comprise a double data rate control circuit 140 that may provide one or more interfaces 142 to, for example, one or more double data rate (DDR) random access memory (RAM) interfaces, a processor interface and/or one or more general purpose input/outputs or interfaces (not shown). Each of the circuits 104 a ″– 104 n ″ may comprise, in one example, a FIFO, a FIFO manager, flow control logic, a media access controller and a memory and statistics support. The circuit 100 ″ may further comprise a number of (e.g., four) SFI-4 controllers 144 , a number of FIFOs 146 , a number of controllers (e.g., XAUI/XGXS) 148 , and a configurable multiplexer circuit 150 that may be configured to select between, for example, an SFI-4, a XAUI and/or a SFI-5 mode. The circuits 106 a ″– 106 n ″ may be implemented, in one example, as quad 3.125 GB/s SERDES circuits. In one example, the circuits 134 , 136 , 138 and 140 may be implemented in the customizable logic block 122 and the circuits 144 , 146 , 148 and 150 may be implemented via the customizable logic blocks 120 a ″– 120 n″.

Referring to FIG. 7 , a block diagram illustrating another example application of the circuit 100 ″ of FIG. 5 is shown. In one example, the interface controller 102 ″ may be configured to couple the circuit 100 ″ to an application specific integrated circuit (ASIC), traffic manager or network processor (NP) chip 152 . The customizable logic block 122 may be configured to provide control of the MACS 104 a ″– 104 n ″ and the logic blocks 120 a ″– 120 n ″ may be configured to provide a reconciliation function between the respective MACs 104 a ″– 104 n ″ and the respective serializer/deserializers 106 a ″– 106 n ″. Each of the serializer/deserializers 106 a ″– 106 n ″ may be configured to support an interface protocol (e.g., XAUI, etc.) for communicating with a respective media dependent interface (MDI) 154 a – 154 n.

In one example, each of the MDIs 154 a – 154 n may be implemented as an electrical-to-optical interface (e.g., a XENPAK compliant 10 Gb/s Ethernet (10 GbE) transceiver). For example, the circuits 154 a – 154 n may be implemented as optical modules incorporating a complete transmit and receive physical layer functionality from a 10 Gb/s optical interface to an XAUI (e.g., four channels X 3.125 Gb/s electrical interface). The circuits 154 a – 154 n may be further configured to provide 8B/10B and/or 64B/66B coding and a MDIO control interface.

In one example, the serializer/deserializers 106 a ″– 106 n ″ may be configured to support one or more different bus protocols (e.g., XAUI, SPI-5, etc.). For example, the serializer/deserializer 106 a ″– 106 n ″ may be implemented with ethernet (e.g., GigaBlaze) and/or optical (e.g., HyperPHY) cores (GigaBlaze and HyperPHY are trademarks of LSI Logic Corp., Milpitas, Calif.). The present invention may reduce the number of pins used for coupling the traffic manager and MAC chips.

Referring to FIG. 8 , a block diagram illustrating a number of examples of ethernet configurations that may be implemented with a preferred embodiment of the present invention. A reference ethernet implementation may include a number of functions and protocols (e.g., column 160 ). For example, a link layer controller may communicate with higher layers using the SPI-4 P2 protocol or some other protocol. A reconciliation layer may communicate with a subsequent layer (or device) via a 10 GB/s medium independent interface (XGMII) protocol. In one example, optional 10 GB/s extender sublayers (XGXS) may be implemented to convert the short run XGMII protocol to a long run 10 GB/s attachment unit interface (XAUI) protocol and back again. An encoding/decoding/error correcting layer may include a physical coding sublayer (PCS), a physical medium attachment (PMA) and support for the XGMII protocol, as well as SFI-4 and XSBI protocols. The encoding/decoding/error correcting layer may communicate with a physical medium dependent interface (PMD) layer that provides the physical network connection via a medium dependent interface (MDI) protocol.

In one example, the higher layers of an ethernet implementation may be provided with an application specific integrated circuit (ASIC), a traffic manager or an NP chip that may be interfaced with the present invention. In general, the present invention may provide a flexible platform that generally supports implementing various ethernet layers that communicate with the higher layers via the link layer in a number of ways. The present invention may provide, for example, a cost effective building block for implementing high speed (e.g., 40 GB/s) optical networks.

Referring to column 162 of FIG. 8 , the reconciliation layer 164 may communicate with a 10 GB/s serializer/deserializer 166 via the XGMII protocol. The serializer/deserializer 166 may communicate with the physical layer via an optical interface 168 . The serializer/deserializer 166 may be configured to manage the error correction operations and physical medium attachment functions.

Referring to column 170 of FIG. 8 , the reconciliation layer 164 may communicate with a serializer/deserializer 172 via an XMGII protocol. The serializer/deserializer 172 may be configured to communicate via an XAUI protocol with, for example, an optical interface 174 (e.g., in a 10G Ethernet Forum 10GBASE-X channel) or an electrical-to-optical interface 176 (e.g., in a 10G Ethernet Forum 10GBASE-R channel). The interfaces 174 and 176 may be configured to provide the error correction, physical medium attachment and physical medium dependent interface functions.

Referring to column 180 of FIG. 8 , an alternative 10G Ethernet Forum 10GBASE-R implementation may comprise a LAN PHY chip (or core) 182 configured to couple (i) to the higher layers via an SPI-4 P2 or other interface and (ii) to a serializer/deserializer 184 via an SFI-4 or XSBI interface. The serializer/deserializer 184 may provide the physical medium attachment interface for coupling to an optical interface 186 . The optical interface 186 generally provides the physical medium dependent interface.

Referring to column 190 of FIG. 8 , a 10G Ethernet Forum 10GBASE-W implementation may comprise a serializer/deserializer 192 , a serializer/deserializer 194 , a physical coding sublayer (PCS) 196 , a single universal physical medium dependent (PMD) interface (SUPI) 198 and an optical link 200 . The serializer/deserializer 192 may connect (i) to the reconciliation layer 164 via the XGMII protocol and (ii) to the serializer/deserializer 194 via an XAUI protocol. The serializer/deserializer 194 may connect to the PCS 196 via the XGMII protocol. The PCS 196 may connect to the SUPI 198 via an SFI-4 or XSBI protocol. The SUPI 198 may provide the interface to the optical link to the physical layer.

Referring to column 210 of FIG. 8 , an alternative implementation of the 10G Ethernet Forum 10GBASE-W channel may comprise a WAN MAC/PHY chip 212 coupled between the higher layers and a SUPI 214 . The SUPI 214 may be configured to couple the channel to the physical layer via an optical link 216 .

Referring to FIG. 9 , a block diagram illustrating an alternative example link layer implementation in accordance with a preferred embodiment of the present invention is shown. The present invention may be configured to implement a 40 GB/s SPI-5 link layer. For example, the link layer may be implemented in an ASIC or traffic manager/NP chip 220 . The circuit 100 may be configured to provide a framer 222 , a forward error correction (FEC) block 224 , and a 40 GB/s serializer/deserializer (SERDES) 226 . The framer 222 may be implemented in (i) dedicated logic, (ii) customizable user defined logic, or (iii) a combination of dedicated and custom logic. The FEC 224 may be implemented in customizable user defined logic (e.g., the custom logic described in connection with FIGS. 3–5 above).

The framer 222 may be coupled (i) to the ASIC 220 via an SPI-5 protocol and (ii) to the FEC 224 via an SFI-5 protocol. The FEC 224 may be coupled to the SERDES 226 via an SFI-5 protocol. The SERDES 226 may be configured to couple the implementation to the physical layer via a laser/optical physical medium dependent interface 228 .

Referring to FIG. 10 , a more detailed block diagram of the example implementation of FIG. 9 is shown. In one example, the circuit 100 ″ (described in more detail in connection with FIGS. 5 and 6 above) may be configured as a 40 GB/s SPI-5 implementation. The circuit 100 ″ may be coupled via an SPI-5 interface to an ASIC or traffic manager/NP chip that may be configured to implement higher network layers of, for example, an OC-768 compliant network.

The circuit 100 ″ may be configured to provide a number of MAC control blocks 230 , a number of MACs 104 a ″– 104 n ″, a number of reconciliation blocks 232 a – 232 n , the framer block 222 , the FEC block 224 , and a number of SERDES 106 a ″– 106 n ″. In one example, the reconciliation blocks 232 a – 232 n , the framer 222 , and the FEC block 224 may be implemented in the custom logic blocks 120 a ′– 120 n ′. The MAC control blocks 230 a – 230 n may be implemented in the custom logic 122 . However other configurations and resource assignments may be implemented to meet the design criteria of a particular application. The circuit 100 ″ may be coupled to the 40 GB/s SERDES 226 via an SFI-5 interface.

While the invention has been particularly shown and described with reference to the preferred embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made without departing from the spirit and scope of the invention.