Bus Training for Interconnected Memory Dice Using Combined Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies face challenges in training multiple interconnected memory dice for a common bus, particularly in scenarios where multi-purpose commands (MPCs) are not available, leading to potential data errors and system malfunctions.
Innovation Solution
The proposed solution involves combining detected bits from multiple memory dice using logical operations such as AND or OR to produce feedback bits, which are then used to train the dice for a common bus timing parameter, allowing for simultaneous training without the need for MPCs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple memory dice are trained individually using existing technologies, then each die can be trained for bus timing, but the training process becomes complex and incompatible with systems that cannot utilize multi-purpose commands (MPCs)
Solution Approach 1:
The patent combines the training responses from multiple memory dice by logically OR-ing their individual responses to form a single composite response. This merging approach allows all dice to be trained simultaneously on the same bus without requiring complex individual training sequences or MPCs, thereby reducing training process complexity while maintaining reliability.
2Adaptability or versatility
If multi-purpose commands (MPCs) are used for training multiple dice, then simultaneous training is enabled, but systems without MPC support cannot perform bus training
Solution Approach 1:
The patent creates a universal bus training method that works across different memory system architectures with or without MPC support. By using standard bus commands and logically combining responses from multiple dice, the training procedure becomes universally applicable to various memory configurations, enhancing adaptability without sacrificing training capability.
3Measurement precision
If sequential training of multiple dice is performed, then each die receives dedicated training attention, but the training time and system downtime increase
Solution Approach 1:
The patent enables continuous bus training by allowing multiple memory dice to be trained simultaneously through their combined response to the same training commands. This eliminates the need to stop training one die before starting another, maintaining continuous useful action and significantly reducing total training duration while preserving training accuracy through logical response combination.
Data Source
AI summary
Described apparatuses and methods facilitate bus training with multiple dice, such as multiple memory dice. A controller can communicate with multiple dice to perform bus training by sending a test pattern and receiving in return a feedback pattern indicative of the bits detected by the dice. Because suitable signal timing can differ between dice, even those using the same bus, a controller may train each die separately from the others. In some situations, however, individualized training may be infeasible. To accommodate such situations, logic associated with two or more dice can combine, using at least one logical operation, bits as detected from the test pattern into a combined feedback pattern. A timing parameter that is jointly suitable for multiple dice can be determined, and the bus training may be concluded, responsive to the combined feedback pattern matching the test pattern. The multiple dice may be stacked or linked.