Abstract:
Methods and apparatus for Ethernet auto-negotiation (AN) with parallel detect for 10G DAC or other non-auto-negotiated modes. AN base pages are transmitted from an Ethernet apparatus to advertise the ability to support at least one Institute of Electrical and Electronics Engineers (IEEE) 802.3 Ethernet specification supporting AN. A receiver and associated processing circuitry is configured to perform two detection modes in parallel, including a first detection mode that looks for a valid signal transmitted from an Ethernet link peer that does not support AN and a second detection mode looking for AN pages from an IEEE 802.3 Ethernet link peer that supports AN. If the link peer does not support AN, an Ethernet link is set up to use signaling in accordance with the Ethernet specification that does not support AN. If the link peer supports AN, an Ethernet link is set up using a corresponding IEEE 802.3 Ethernet link supporting AN. Supported non-AN Ethernet links include 10G DAC links.
Abstract:
Technologies for capabilities exchange include a network port logic having a communication link coupled to a remote link partner. The port logic transmits local host loss information to the link partner and receives remote host loss information from the link partner. The port logic may communicate the host loss information via an autonegotiation base page, an autonegotiation next page, or a PMD control frame. The port logic determines total channel loss based on the local host loss, the remote host loss, and cable loss. The port logic may bring the communication link up without forward error correction (FEC) if the total channel loss is less than a FEC limit, may bring the link up with FEC if the total loss is less than a specification limit, or may not bring the link up if the total channel loss is above the specification limit. Other embodiments are described and claimed.
Abstract:
Technologies for high-speed data transmission include a network port logic having one or more communication lanes coupled to a forward error correction (FEC) sublayer and a physical coding sublayer (PCS). To transmit data, the PCS encodes the data to be transmitted into encoded data blocks using a 66b/64b line code and inserts alignment marker blocks after every 16,383 encoded data blocks. The FEC encodes the encoded data blocks into 80-block FEC codewords starting at a predefined offset from an alignment marker. Thus, each alignment marker is at one of five predefined offsets from the beginning of an FEC codeword. Each alignment marker may include a unique block type field usable with FEC encoding. The PCS may include one or more logical lanes, each operating at 25 Gb/s. Embodiments of the network port logic may include a single PCS lane or sixteen PCS lanes. Other embodiments are described and claimed.
Abstract:
Technologies for transmitter equalization in a communication system include reading local transmitter equalization settings from a transmitter equalization register of a first communication device and writing the local transmitter equalization settings to a transmitter equalization register of a second communication device communicatively coupled with the first communication device via a chip-to-chip communication link. Additionally, requested transmitter equalization settings may be read from the transmitter equalization register of the second communication device and written to the transmitter equalization register of the first communication device. The reading and writing process may be repeated for the opposite communication direction and for other communication lane interfaces of the first and second communication devices.
Abstract:
Methods, apparatus and systems for measuring signal transition times for a four-level pulse modulated amplitude (PAM4) transmitter. During a test procedure, a PAM4 transmitter is configured to repetitively transmitting a four-level test pattern, which is captured and digitized. The digitized data is processed to generate a linear-fitted waveform. A voltage modulation amplitude (VMA) level for each of a −1 and +1 PAM4 signal level is measured and used to derive 20% and 80% VMA levels in an eye diagram. The rise transition time is then determined by measuring the time interval between when a rising signal crosses the 20% and 80% VMA levels, and the fall transition time is determined by measuring the time interval between when a falling signal crosses the 80% and 20% VMA levels.
Abstract:
Methods, apparatus and systems for implementing for implementing Energy-Efficient Ethernet (EEE) refresh and wake signaling for high-speed Ethernet links. During an EEE refresh or wake signaling period, ALERT frames are transmitted between first and second Ethernet interfaces on opposing ends of the link, with a first ALERT frame being sent from a first Ethernet interface and a second ALERT frame being returned from the second Ethernet interface. The ALERT frames have a length that is different that the length of Physical Media Attachment (PMA) frames, and the returned ALERT frames include frame alignment offset data identifying a relative offset between an ALERT frame and a predetermined location in a PMA frame, and countdown data. The frame alignment offset data and countdown data are employed to facilitate a rapid transition from the link training mode to the data mode.
Abstract:
Technologies for robust data transmission include a network port logic having a physical coding sublayer (PCS). The PCS may transmit a series of rapid alignment markers (RAMs) to a link partner, with each RAM indicative of a counter value. The PCS transitions to a sleep state if the counter value equals two and a low power idle (LPI) command is set by an upper-layer client. The PCS transitions to an active state if the counter value equals one and the LPI command is not set. The PCS may receive a low power idle symbol (LI) from the link partner and start a guard timer in response to receipt of the LI symbol. The PCS transitions to a sleep state if the guard timer expires and transitions to the active state if data other than LI is received prior to expiration of the guard timer. Other embodiments are described and claimed.
Abstract:
Embodiments of the present disclosure provide configurations for testing arrangements for testing multi-lane active cables. In one embodiment, a testing arrangement may comprise a testing module comprising a pattern generator to be coupled with an active cable having a plurality of lanes to generate a test pattern to be transmitted over the active cable, wherein the test pattern is to be transmitted at least over two or more lanes of the active cable that are concatenated, and a processing unit to be coupled with the active cable to process a result of the transmission of the test pattern over the active cable. The arrangement may further include a plurality of testing cables to concatenate two or more of the lanes of the active cable, to enable the transmission of the test pattern over the concatenated lanes of the active cable. Other embodiments may be described and/or claimed.
Abstract:
Technologies for transmitter equalization in a communication system include reading local transmitter equalization settings from a transmitter equalization register of a first communication device and writing the local transmitter equalization settings to a transmitter equalization register of a second communication device communicatively coupled with the first communication device via a chip-to-chip communication link. Additionally, requested transmitter equalization settings may be read from the transmitter equalization register of the second communication device and written to the transmitter equalization register of the first communication device. The reading and writing process may be repeated for the opposite communication direction and for other communication lane interfaces of the first and second communication devices.
Abstract:
Methods, apparatus, and systems for preventing false packet acceptance in high-speed links. Under one aspect, correctable symbol errors are detected, and determination is made to whether a symbol error rate or ratio (SER) exceeds an SER threshold. In response to detection of such a condition, the link is disconnected or temporarily paused. The value for the SER threshold is determined using a statistical analysis of various link parameters to meet desired performance levels, such as a mean time to false packet acceptance (MTTFPA) of >approximately 15 billion years while providing a mean time to disconnect of >100 years.