Overlay Code Retrieval From Host Memory to Cut Latency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Retrieving code from non-executable memory in memory systems is time-consuming and can increase latency and wearout, consuming resources that could be used for other memory access operations.
Innovation Solution
Retrieve code from a host system with lower access latency instead of non-executable memory to reduce retrieval latency and conserve resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If code is retrieved from non-executable memory, then code storage capacity is sufficient, but retrieval latency increases and memory wearout accelerates
Solution Approach 1:
The patent segments code into two categories: frequently accessed code stored in high-speed executable memory and less frequently accessed code stored in non-executable memory. This segmentation allows the system to retrieve code from the appropriate storage location based on access patterns, reducing latency for frequent operations while maintaining sufficient total storage capacity.
Solution Approach 2:
The patent implements preliminary action by pre-loading frequently accessed code into executable memory before it is needed. The system monitors code access patterns and proactively transfers code from non-executable memory to executable memory in advance, ensuring that high-priority code is already in fast storage when needed, thereby reducing retrieval latency.
2Quantity of substance
If code is retrieved from non-executable memory, then code storage capacity is sufficient, but memory wearout increases
Solution Approach 1:
The patent segments code access into two paths: frequently accessed code is served from executable memory, while less frequently accessed code resides in non-executable memory. This segmentation reduces the number of read operations on non-executable memory, thereby reducing wearout and extending its operational lifespan while maintaining sufficient total storage capacity.
Solution Approach 2:
The patent creates copies of frequently accessed code in executable memory from the master copy in non-executable memory. By serving code requests from the executable memory copy rather than repeatedly reading from non-executable memory, the system reduces wear on the non-executable memory while maintaining adequate storage capacity through the original master copy.
3Quantity of substance
If code is retrieved from non-executable memory, then all code can be stored, but resource utilization for other memory operations decreases
Solution Approach 1:
The patent segments memory resources into executable memory for frequent code access and non-executable memory for less frequent access. This segmentation enables parallel operations where the memory system can service code requests from executable memory while simultaneously handling other memory operations involving non-executable memory, thereby improving overall resource utilization while maintaining sufficient total storage capacity.
Solution Approach 2:
The patent implements preliminary action by pre-transferring frequently accessed code to executable memory before peak demand occurs. This allows the memory system to utilize non-executable memory for other operations during low-demand periods while ensuring fast code availability during high-demand periods, thereby improving overall resource utilization across different time periods.
Data Source
AI summary
Methods, systems, and devices for overlay code retrieval from a host system are described. A memory system may determine that a set of code for execution by a processor of the memory system is absent from an executable memory of the processor. The memory system may prevent the processor from retrieving the set of code from a non-volatile memory of the memory system based on the set of code being designated for retrieval from a host system. The memory system may retrieve the set of code from a memory of a host system, instead of retrieving the set of code from the non-volatile memory, based on the set of code being designated for retrieval from the host system.