DDR Memory Vref Training 2D Voltage Domain Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing DDR memory data eye training methods are inefficient and unreliable, particularly in systems with varying hardware platforms, as they fail to adequately address Vref tolerances and data transfer speed issues, limiting system boot reliability and complexity.
Innovation Solution
A method for DDR memory reference voltage training in the voltage domain, which involves a two-dimensional analysis combining time and voltage domains to optimize timing and Vref margins, allowing for adaptable Vref settings on a byte lane, rank, or channel basis, enabling robust data eye formation across different hardware configurations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If time domain training is used with nominal Vref value setting, then training process is simplified, but Vref tolerances cannot be adequately met and system boot reliability is reduced
Solution Approach 1:
The patent transitions from one-dimensional time domain training to two-dimensional training by adding voltage domain analysis. The method performs training in both time domain (data strobe timing) and voltage domain (Vref values), creating a comprehensive 2D training matrix that evaluates multiple Vref levels simultaneously with timing adjustments. This dimensional expansion enables adequate Vref tolerance coverage while maintaining system boot reliability across diverse hardware platforms.
Solution Approach 2:
The patent systematically varies multiple parameters including Vref voltage levels, data strobe timing values, and their combinations to create a comprehensive training matrix. By changing these parameters across different dimensions and evaluating their effects, the method identifies optimal settings that satisfy Vref tolerances and ensure reliable system boot across varying hardware configurations.
2Device complexity
If time domain training is used, then training algorithm is simplified, but Vref tolerances and data transfer speed optimization are insufficient
Solution Approach 1:
The patent adds voltage domain analysis to the existing time domain training approach, creating a two-dimensional training matrix. This matrix evaluates multiple Vref levels (e.g., nominal Vref and adjusted Vref values) combined with different data strobe timing values. By performing training in both dimensions simultaneously, the method comprehensively addresses Vref tolerances and identifies optimal timing-voltage combinations that maximize data transfer speed while meeting precision requirements.
3Ease of operation
If fixed Vref setting is used, then system operation is simplified, but adaptability to different hardware platforms is reduced
Solution Approach 1:
The patent implements dynamic Vref adjustment by performing training at multiple Vref levels and selecting the optimal setting based on measured performance metrics. The training process evaluates different Vref values (nominal and adjusted) and determines the best configuration for the specific hardware platform. This dynamic approach allows the system to automatically adapt to varying hardware conditions while maintaining simplified operation through automated training procedures.
Solution Approach 2:
The patent systematically varies Vref parameters and timing parameters to create multiple training configurations. By changing these parameters and measuring system response, the method identifies optimal settings that adapt to different hardware platforms. The training algorithm automatically selects the best Vref and timing combination, providing hardware-specific optimization without requiring manual intervention.
Data Source
AI summary
A method is provided for performing memory operations in response to instructions to perform a double data rate (DDR) memory reference voltage training in the voltage domain by a processing device and determining a DDR memory reference voltage and a DDR memory delay time based upon the memory operation. Computer readable storage media are also provided. A circuit is provided that includes a communication interface portion coupled to a memory and to a processing device. The circuit also includes a circuit portion, coupled to the communication interface portion that has a hardware state machine or an algorithm. The state machine or algorithm provides instructions to the processing device to perform a double data rate (DDR) reference voltage training in the voltage domain.


