Abstract:
Techniques for reducing idle power consumption of a port are described herein. An example method includes determining device presence using a pull-down resistor disposed in a downstream port. The method also includes initiating a low power state of a link between the downstream port and an upstream device. The method also includes disabling the pull-down resistor in response to initiating the low power state.
Abstract:
A system and method is described for simplifying implementation of repeater (e.g., re-driver/re-timer) module implementation in high-data-rate interconnects that carry a relatively low-data-rate clock signal as well as the data stream (e.g., PCIe). At the endpoint, any information critical to the function of the repeater (e.g., the most recent data rate negotiated by a pair of endpoints communicating through the repeater) is embedded in the clock signal by pulse-width modulation as ordered sets. The repeater only needs to read the clock-embedded information rather than decoding the data stream. Thus repeaters for such applications reconstruct the high-rate data-stream while actually decoding only the low-rate clock signal. Because the clock-signal protocol is independent of the data-stream protocol, the repeater's operation is protocol-agnostic with respect to the data-stream.
Abstract:
A system and method is described for simplifying implementation of repeater (e.g., re-driver/re-timer) module implementation in high-data-rate interconnects that carry a relatively low-data-rate clock signal as well as the data stream (e.g., PCIe). At the endpoint, any information critical to the function of the repeater (e.g., the most recent data rate negotiated by a pair of endpoints communicating through the repeater) is embedded in the clock signal by pulse-width modulation as ordered sets. The repeater only needs to read the clock-embedded information rather than decoding the data stream. Thus repeaters for such applications reconstruct the high-rate data-stream while actually decoding only the low-rate clock signal. Because the clock-signal protocol is independent of the data-stream protocol, the repeater's operation is protocol-agnostic with respect to the data-stream.
Abstract:
A system and method is described for simplifying implementation of repeater (e.g., re-driver/re-timer) module implementation in high-data-rate interconnects that carry a relatively low-data-rate clock signal as well as the data stream (e.g., PCIe). At the endpoint, any information critical to the function of the repeater (e.g., the most recent data rate negotiated by a pair of endpoints communicating through the repeater) is embedded in the clock signal by pulse-width modulation as ordered sets. The repeater only needs to read the clock-embedded information rather than decoding the data stream. Thus repeaters for such applications reconstruct the high-rate data-stream while actually decoding only the low-rate clock signal. Because the clock-signal protocol is independent of the data-stream protocol, the repeater's operation is protocol-agnostic with respect to the data-stream.
Abstract:
A first state of an interconnect protocol is entered. A particular signal is sent according to the protocol to a device over a link. During the first state, it is detected that a response to the particular signal is received in the first state. It is determined that the device supports a configuration mode outside the protocol based on the received response. The configuration mode is entered based on the response. One or more in-band configuration messages are sent within the configuration mode.
Abstract:
A method and system for communicating data between two devices are described herein. The method detects an electrical signal of a first protocol from a first device in a repeater, wherein the first protocol comprises single-ended signaling. The method also determines the speed of the electrical signal. Additionally, the method converts the electrical signal of the first protocol into an electrical signal of a second protocol based on the speed of the electrical signal. The second protocol comprises differential signaling. Furthermore, the method sends the electrical signal of the second protocol to a second device. In addition, the method stops the electrical signal of the second protocol to the second device when the electrical signal of the second protocol indicates an end of data flow.
Abstract:
Systems and methods of operating a serial interconnect interface provide for generating a pulse in response to a state change in a data signal of the serial interface interconnect, and transmitting the pulse from a physical layer of the serial interconnect interface to a link layer of the serial interconnect interface. The duration of the pulse can be selected based on whether the state change corresponds to an end of packet (EOP) condition. In addition, the data signal may include a non return to zero invert (NRZI) encoded signal, wherein the pulse is part of a digital NRZI signal.
Abstract:
Systems and methods for operating a universal serial bus are described herein. The method includes sending packet data from a USB2 device to a USB2 host on a pair of signal lines, and after sending the packet data, sending an End-Of-Packet (EOP) signal from the USB2 device to the USB2 host. The method also includes, entering the USB2 device into idle state after sending the EOP signal. The method also includes sending a digital ping from the USB2 device to the USB2 host to indicate device presence during idle state.
Abstract:
Techniques for embedded high speed serial interface methods are described herein. The method includes issuing a single-ended one (SE1) signal on each of a pair of embedded high speed serial interface data lines, the SE1 indicating a register access protocol (RAP) message follows the SE1 signal. The method also includes accessing a register of an embedded high speed serial interface component based on the RAP message.
Abstract:
Systems and methods for detecting Input/Output (I/O) device connection are described herein. The method includes physically coupling an I/O device to a host port through a first signal line and a second signal line. The method also includes driving the first signal line or the second signal line high via an active buffer of the I/O device. The method also includes providing an acknowledgement signal from the host to the device through the other signal line that is not being driven high by the active buffer of the I/O device.