Chip Select Signal Training for Memory Devices
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In memory systems, the varying propagation delays and clock signal differences among memory devices lead to misinterpretation of chip select signals, causing some memory devices to malfunction during CS signal read operations, especially in multi-rank DIMMs where accurate timing is crucial for data access.
Innovation Solution
The implementation of a method to determine optimal timing offsets and reference voltage values for chip select signals, allowing memory devices to accurately read active and inactive parts of the signal by using a composite eye calculation based on feedback from multiple devices, ensuring all devices in a rank can correctly interpret the signal.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If memory devices operate at higher frequencies to increase data access speed, then productivity is improved, but timing synchronization among multiple memory devices deteriorates due to varying propagation delays
Solution Approach 1:
The patent applies preliminary action by performing timing offset calibration and composite eye calculation during the training phase before actual data operations. The memory controller pre-determines the optimal timing offsets for each memory device by analyzing composite eye diagrams, ensuring that all devices are properly synchronized before high-frequency data access begins. This preliminary timing adjustment prevents synchronization issues during actual operation.
2Quantity of substance
If multiple memory devices are coupled to the same bus to increase memory capacity, then quantity of substance is improved, but signal interpretation accuracy deteriorates due to varying propagation delays and clock signal differences
Solution Approach 1:
The patent applies local quality by allowing each memory device to have its own device-specific timing offset value that compensates for its individual propagation delay characteristics. Instead of using a uniform timing reference for all devices, the memory controller determines and applies customized timing offsets for each device based on its specific electrical characteristics and position on the bus, thereby maintaining signal interpretation accuracy across multiple devices.
Solution Approach 2:
The patent applies parameter changes by dynamically adjusting the timing offset parameter for each memory device based on composite eye diagram analysis. The memory controller varies the timing offset values during training to optimize the sampling point for each device, selecting the parameter values that maximize signal interpretation accuracy for each individual device in the multi-device system.
3Measurement precision
If timing offsets are adjusted to improve chip select signal reading accuracy, then measurement precision is improved, but device complexity increases due to additional training procedures
Solution Approach 1:
The patent applies merging by combining the timing offset calibration process with the existing memory training procedures. The composite eye calculation and timing offset determination are integrated into the standard memory initialization sequence, so that the same training infrastructure and control logic are used for both traditional training tasks and the new timing calibration function, thereby minimizing the increase in device complexity.
Data Source
AI summary
A reference voltage value and a chip select (CS) signal timing delay provided to memory devices can be determined based on samples of the CS signal received by the memory devices. The CS signal can be provided to the memory devices with varying time delays and for various reference voltages. Various samples of the CS signal from the memory devices can indicate different times for rising and falling edges of the CS signal. A composite signal eye can be generated by the latest occurring rising edge and the earliest occurring falling edge of the CS signal. The reference voltage value and timing delay can be chosen based on the composite signal eye width that is the closest to a reference eye width.