Dynamic Private Page Pools for Compressed Data Footprint Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing volume of sensor data captured by modern computing devices often leads to memory overload and inefficient use of memory buffers due to unpredictable compression ratios, resulting in unused memory space and reduced space-saving benefits from bandwidth compression methods.
Innovation Solution
A private page pool is allocated separately from main memory to dynamically manage memory pages based on demand and post-compression, allowing for quick allocation, reallocation, and freeing of memory pages without relying on main memory management processes, thereby reducing memory waste and address translation overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If memory buffers are allocated to retain sensor data before compression, then data can be captured and compressed, but memory space is wasted because the allocation is based on uncompressed size and cannot be re-allocated
Solution Approach 1:
The patent implements dynamic memory allocation by creating a page pool that can adaptively allocate and deallocate memory pages based on actual compression ratios. Instead of static allocation, the system continuously monitors compression performance and adjusts memory usage accordingly, allowing the same memory to serve multiple compression tasks over time.
Solution Approach 2:
The system changes the parameter of memory allocation from fixed size to variable size based on compression ratio. By monitoring actual compression performance and adjusting the number of memory pages allocated accordingly, the system optimizes memory utilization while ensuring sufficient space for data retention during compression operations.
2Productivity
If large memory buffers are allocated for high resolution sensor data, then data capture is enabled, but memory overload occurs and other processes are affected
Solution Approach 1:
The patent implements dynamic memory allocation by creating a page pool that can adaptively allocate and deallocate memory pages based on actual compression ratios. Instead of static allocation, the system continuously monitors compression performance and adjusts memory usage accordingly, allowing the same memory to serve multiple compression tasks over time.
Solution Approach 2:
The system recovers and reuses memory pages after compression operations complete. The page pool mechanism collects freed memory pages and makes them available for subsequent compression tasks, effectively recycling memory resources rather than leaving them permanently allocated or wasted.
3Ease of operation
If memory is allocated before knowing the compression ratio, then data can be captured, but the space saving benefits of compression are lost
Solution Approach 1:
The patent applies preliminary action by pre-allocating a page pool with sufficient memory pages to handle maximum expected data volume. This preliminary allocation ensures data capture can proceed without interruption, while the system then optimizes actual usage by dealinglocate unused pages based on real compression ratios, maintaining both operational simplicity and compression efficiency.
Solution Approach 2:
The system implements feedback by monitoring actual compression ratios and using this information to adjust memory allocation in the page pool. The feedback loop ensures that memory usage aligns with actual compression performance, preventing both waste from over-allocation and insufficiency from under-allocation.
Data Source
AI summary
A method and apparatus for reducing a memory footprint of compressed data by dynamically allocating and recycling memory pages. The method may include allocating a dedicated subset of the memory pages to a private page pool separate from a main memory such that the dedicated subset of the memory pages is allocatable separately from the main memory, receiving compressed data, allocating at least one physical address of the dedicated subset of the memory pages from the private page pool for the compressed data, and writing the compressed data to the one or more physical addresses.