Hardware Transmit Equalization for High Speed Links
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Preparing point-to-point links for high-speed data transfer is a slow and complex process in modern computing systems, which hampers efficient communication between devices.
Innovation Solution
The method involves subdividing communication links into two subsets, where one subset operates at a target high speed for equalization training, and the other at a lower speed to convey feedback, allowing both subsets to eventually operate at the target speed for data transfer.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If traditional equalization training is performed on all lanes at high speed, then data transfer rate is improved, but training time and complexity increase significantly
Solution Approach 1:
The patent divides the communication link into two distinct subsets: a first subset of lanes used for high-speed data transfer and a second subset of lanes used for feedback communication. This segmentation allows equalization training to be performed on the first subset at high speed while the second subset operates at lower speeds to convey training feedback, thereby reducing overall training time while maintaining high data transfer rates.
2Productivity
If equalization training is performed on all lanes simultaneously at high speed, then data transfer efficiency is improved, but system complexity and difficulty of implementation increase
Solution Approach 1:
The communication link is segmented into two functional subsets: first subset lanes dedicated to high-speed data transfer and second subset lanes dedicated to feedback communication. This segmentation simplifies the equalization training process by separating the high-speed data path from the feedback path, making the system easier to implement and manage while maintaining high productivity.
Solution Approach 2:
The second subset of lanes acts as an intermediary channel for conveying feedback information during equalization training. This intermediary feedback path allows the transmitter to adjust equalization settings based on receiver feedback without requiring the entire link to operate at high speed during training, thereby reducing complexity while maintaining data transfer efficiency.
3Measurement precision
If feedback is conveyed during equalization training, then training accuracy is improved, but the lanes used for feedback must operate at lower speeds increasing overall training duration
Solution Approach 1:
The patent segments the communication link into first subset lanes for high-speed operation and second subset lanes for feedback communication. This allows the feedback lanes to operate at lower speeds during training without impacting the high-speed performance of the data transfer lanes, maintaining training accuracy while preserving overall system speed capability.
Solution Approach 2:
Different subsets of lanes are assigned different operational characteristics: the first subset is optimized for high-speed data transfer while the second subset is optimized for reliable feedback communication at lower speeds. This local differentiation allows each subset to operate at optimal speeds for its specific function, maintaining training accuracy without compromising overall system performance.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Systems, apparatuses, and methods for performing transmit equalization at a target high speed are disclosed. A computing system includes at least a transmitter, receiver, and a communication channel connecting the transmitter and the receiver. The communication channel includes a plurality of lanes which are subdivided into a first subset of lanes and a second subset of lanes. During equalization training, the first subset of lanes operate at a first speed while the second subset of lanes operate at a second speed. The first speed is the desired target speed for operating the communication link while the second speed is a relatively low speed capable of reliably carrying data over a given lane prior to equalization training. The first subset of lanes are trained at the first speed while feedback is conveyed from the receiver to the transmitter using the second subset of lanes operating at the second speed.