# US 2012/0250698 A1 · Large Interconnect Fabrics

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2012/0250698 A1
- Document ID: US20120250698A1
- Status: Published application
- Kind: A1
- Application number: 13/417,091
- Filed: 2012-03-09
- Published: 2012-10-04
- Earliest claimed: 2011-03-10
- Assignees: Cisco Technology, Inc.
- Inventors: Madian Somasundaram
- Web page: https://patentgrep.com/patent/US20120250698A1

## CPC

- H04L 49/3027: {Output queuing}
- H04L 49/101: using crossbar or matrix
- H04L 49/3045: {Virtual queuing}
- H04L 49/508: {Head of Line Blocking Avoidance}
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)

## Abstract

In one embodiment, a network switch includes multiple chips communicably coupled together and a buffered crossbar. Each chip is coupled to every other chip with two bi-directional serial channels and includes a slice of the buffered crossbar. One or more input ports, one or more output ports, and an input logic module are coupled to the plurality of chips, and the input logic module is configured to receive a packet of data, allocate the packet of data into one or more data fragments, and distribute the packet of data to the buffered crossbar. An output logic module is coupled to the chips and configured to retrieve the packet of data from the buffered crossbar, reconstruct the packet of data from the data fragments according to a gather scheme, and transmit the packet of data.

## Claims

### Claim 1

A network switch, comprising:
a plurality of chips communicably coupled together and comprising a buffered crossbar, each chip coupled to each other chip with two bi-directional serial channels, each chip comprising a slice of the buffered crossbar;
one or more input ports;
one or more output ports;
an input logic module coupled to the plurality of chips, the input logic module configured to:
receive a packet of data;
allocate the packet of data into one or more data fragments; and
distribute the packet of data to the buffered crossbar, each data fragment being distributed to a particular chip comprising a slice of the buffered crossbar according to a scatter scheme; and
an output logic module coupled to the plurality of chips, the output logic module configured to:
retrieve the packet of data from the buffered crossbar by retrieving the one or more data fragments from the plurality of chips;
reconstruct the packet of data from the one or more data fragments according to a gather scheme; and
transmit the packet of data.

### Claim 2 (depends on 1)

The network switch of claim 1, wherein the slice of the buffered crossbar comprises
N×N horizontal lines;
N×N vertical lines where N is an integer greater than one; and
a memory at each intersection of the horizontal and vertical lines.

### Claim 3 (depends on 1)

The network switch of claim 1, wherein a portion of the buffered crossbar is allocated in accordance with quality of service.

### Claim 4 (depends on 3)

The network switch of claim 3, the portion of the buffered crossbar is allocated for one of unicast or multicast.

### Claim 5 (depends on 1)

The network switch of claim 1, wherein each of one or more data fragments are about equally sized.

### Claim 6 (depends on 1)

The network switch of claim 1, wherein each of the one or more data fragments are distributed to a particular memory location on the particular chip, the particular memory location corresponding to an intersection of the input port and an intended output port.

### Claim 7 (depends on 1)

The network switch of claim 1, wherein the bi-directional serial channels are time-sliced for each input port and output port.

### Claim 8 (depends on 7)

The network switch of claim 7, wherein the each data fragment being distributed to the particular chip is distributed according to a time slot assigned to a particular input port.

### Claim 9

A system comprising:
a network switch, comprising:
a plurality of chips communicably coupled together and comprising a buffered crossbar, each chip coupled to each other chip with two bi-directional serial channels, each chip comprising a slice of the buffered crossbar;
one or more input ports;
one or more output ports;
an input logic module coupled to the plurality of chips, the input logic module configured to:
receive a packet of data;
allocate the packet of data into one or more data fragments; and
distribute the packet of data to the buffered crossbar, each data fragment being distributed to a particular chip comprising a slice of the buffered crossbar according to a scatter scheme; and
an output logic module coupled to the plurality of chips, the output logic module configured to:
retrieve the packet of data from the buffered crossbar by retrieving the one or more data fragments from the plurality of chips;
reconstruct the packet of data from the one or more data fragments according to a gather scheme; and
transmit the packet of data;
a memory comprising instructions executable by one or more processors; and
the one or more processors coupled to the memory.

### Claim 10 (depends on 9)

The system of claim 9, wherein the slice of the buffered crossbar comprises
N×N horizontal lines;
N×N vertical lines where N is an integer greater than one; and
a memory at each intersection of the horizontal and vertical lines.

### Claim 11 (depends on 9)

The system of claim 9, wherein a portion of the buffered crossbar is allocated in accordance with quality of service.

### Claim 12 (depends on 11)

The system of claim 11, the portion of the buffered crossbar is allocated for one of unicast or multicast.

### Claim 13 (depends on 9)

The system of claim 9, wherein each of one or more data fragments are about equally sized.

### Claim 14 (depends on 9)

The system of claim 9, wherein each of the one or more data fragments are distributed to a particular memory location on the particular chip, the particular memory location corresponding to an intersection of the input port and an intended output port.

### Claim 15 (depends on 9)

The system of claim 9, wherein the bi-directional serial channels are time-sliced for each input port and output port.

### Claim 16 (depends on 15)

The system of claim 15, wherein the each data fragment being distributed to the particular chip is distributed according to a time slot assigned to a particular input port.

### Claim 17

A method comprising:
by one or more network switches:
receive a packet of data at an input logic module coupled to a plurality of chips comprising a buffered crossbar, each chip coupled to each other chip and comprising a slice of the buffered crossbar;
allocate the packet of data into one or more data fragments;
distribute the packet of data to the buffered crossbar through one or more input ports, each data fragment being distributed to a particular chip comprising a slice of the buffered crossbar according to a scatter scheme;
retrieve the packet of data from the buffered crossbar at an output logic module through an output port by retrieving the one or more data fragments from the plurality of chips, the output logic module coupled to the plurality of chips;
reconstruct the packet of data from the one or more data fragments according to a gather scheme; and
transmit the packet of data.

### Claim 18 (depends on 17)

The method of claim 17, wherein each of the one or more data fragments are distributed to a particular memory location on the particular chip, the particular memory location corresponding to an intersection of the input port and an intended output port.
