Memory Receiver Training for Per-Symbol Voltage and Timing Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing memory systems face challenges in efficiently communicating data between host devices and memory devices due to variations in signal characteristics across different unit intervals and symbols, leading to errors in transmission.
Innovation Solution
The implementation of a training procedure that tunes receivers by adjusting reference voltages and sampling times on an individual basis, using baseline training operations and individual training operations to compensate for variations, allowing for independent configuration of reference voltages and sampling times across different unit intervals and symbols.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If individual training operations are performed for each reference voltage and sampling time, then measurement precision and reliability are improved, but device complexity and training time increase
Solution Approach 1:
The training procedure is segmented into distinct phases: baseline training operation that establishes initial reference voltages and sampling times, followed by individual training operations for each unit interval. This segmentation allows systematic adjustment of parameters without overwhelming complexity, addressing the contradiction by breaking down the complex training process into manageable stages.
Solution Approach 2:
The baseline training operation performs preliminary adjustments to establish initial reference voltages and sampling times before individual unit interval training. This preliminary action reduces the complexity of subsequent individual operations by providing a solid foundation, thereby improving measurement precision without proportionally increasing overall device complexity.
2Reliability
If individual training operations are performed for each reference voltage and sampling time, then data transmission reliability is improved, but training time increases
Solution Approach 1:
The training process is divided into baseline training (affecting all unit intervals) and individual unit interval training. This segmentation enables parallel processing potential and systematic progression, improving data transmission reliability through precise individual adjustments while managing training time through structured phases.
Solution Approach 2:
Baseline training operations are performed first to establish initial reference voltages and sampling times that apply to all unit intervals. This preliminary action reduces the total training time required for individual operations by pre-configuring common parameters, thereby improving reliability without proportionally increasing training time.
3Productivity
If baseline training operations affect all unit intervals, then productivity is improved through efficient bulk configuration, but measurement precision may be compromised due to lack of individual optimization
Solution Approach 1:
The training system is segmented into two levels: baseline training that efficiently configures all unit intervals with common parameters (improving productivity), and individual unit interval training that optimizes specific parameters for each interval (improving measurement precision). This dual-level segmentation resolves the contradiction by operating at both bulk and individual levels.
Solution Approach 2:
The baseline training operation serves as a universal configuration that applies to all unit intervals, establishing common reference voltages and sampling times efficiently. This universal approach improves productivity while the subsequent individual training operations provide the necessary precision adjustments, combining the benefits of both bulk and individual optimization.
Data Source
AI summary
Systems, apparatuses, and methods for training procedures on reference voltages and sampling times associated with symbols communicated with a memory device are described. The training procedures may be configured to compensate for variations that may occur in different symbols of a signal. For example, an individual training operation may be performed for each reference voltage within a first unit interval. These individual training operations may allow a reference voltage of the first unit interval to be positionable independent of other reference voltages in the same unit interval or in different unit intervals. In another example, an individual training operation may be performed for the sampling time associated with a reference voltage. These individual training operations may allow a sampling time associated with a reference voltage in the first unit interval to be positionable independent of other sampling times in the same unit interval or in different unit intervals.