Interleaved Memory Transaction Tracking for Unrelated-Length Streams
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing memory systems face challenges in efficiently managing and storing compressed data with varying sizes and formats, particularly when multiple interleaved data streams with unknown lengths are involved, leading to inefficiencies in bandwidth, memory usage, and power consumption.
Innovation Solution
A method for allocating memory space based on the known total number of transactions for interleaved data series, where each series is stored sequentially from opposite ends of the allocated space, reducing the need for tracking registers and minimizing communication overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If traditional memory tracking methods are used for interleaved data streams with unknown lengths, then data can be stored in memory, but the number of tracking registers increases and bandwidth efficiency decreases
Solution Approach 1:
The patent segments the memory space into distinct regions for different data streams, allowing each stream to be tracked independently without requiring complex global tracking mechanisms. This segmentation enables simpler register usage while maintaining efficient memory utilization for interleaved data streams with unknown lengths.
Solution Approach 2:
The patent employs preliminary action by pre-allocating memory regions and establishing tracking mechanisms before data streams are fully received. This allows the system to prepare memory structures in advance, reducing the need for complex real-time tracking registers while maintaining efficient memory usage throughout data reception.
2Reliability
If comprehensive tracking of interleaved data streams is implemented, then data integrity is maintained, but power consumption increases
Solution Approach 1:
The patent applies partial action by implementing tracking mechanisms only for the essential aspects of data stream management rather than comprehensive tracking. This selective tracking maintains sufficient data integrity while significantly reducing the power consumption associated with full-scale tracking of all interleaved streams.
Solution Approach 2:
The patent enables self-service by allowing data streams to be tracked and managed through their inherent properties and memory allocation mechanisms, rather than requiring intensive external tracking resources. This self-managing approach maintains data integrity while minimizing additional power consumption from complex tracking systems.
3Adaptability or versatility
If memory space is allocated without knowing total transaction count, then flexibility is improved, but memory efficiency decreases
Solution Approach 1:
The patent implements dynamics by creating a flexible memory allocation system that can adapt to varying data stream lengths while maintaining efficiency. The system dynamically adjusts memory region assignments and tracking mechanisms based on actual data stream characteristics, allowing flexibility without sacrificing memory efficiency through overly conservative allocations.
Data Source
AI summary
A first series of memory access transactions are received from a first set of compressed data having a number of first transactions indeterminate in a transaction receiver, and a second series of memory access transactions are received from a second set of compressed data having a number of second transactions indeterminate in the transaction receiver. A count of outstanding transactions for the combined first and second sets of compressed data is received in the transaction receiver, and memory space is allocated for the first and second series of transactions based at least in part on the received count of outstanding combined transactions. The first series of transactions are stored sequentially from a first end of the allocated memory space and the second series of transactions are stored sequentially from a second end of the allocated memory space opposite the first end.