Buffer Memory Segmentation for Charged Particle Beam Writing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In charged particle beam writing apparatuses, memory fragmentation and swapping out occur due to increased data processing demands, leading to prolonged data read-out times and potential stops in writing processing, especially when memory control is managed by commercial operating systems.
Innovation Solution
A buffer memory system is implemented with a large and small region size configuration, where large data files are prioritized in the large region and small files in the small region, allowing for efficient data storage and processing across multiple data processing regions, managed independently of commercial OS control.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If memory control is managed by commercial operating systems, then ease of operation is improved, but memory fragmentation and swapping out occur leading to prolonged data read-out times
Solution Approach 1:
The buffer memory is divided into multiple regions (first region, second region, third region) with different characteristics. Each region serves specific purposes: the first region stores writing data for current processing, the second region stores data for next processing regions, and the third region serves as a work area. This segmentation prevents fragmentation by dedicating specific regions for specific functions, thereby maintaining fast data read-out times while preserving ease of operation through structured memory management.
2Productivity
If memory capacity is increased to handle increased data processing demands, then productivity is improved, but memory fragmentation and swapping out occur causing data processing stops
Solution Approach 1:
The system performs preliminary actions by pre-loading writing data for next data processing regions into the second region before the actual writing process. The control unit predicts future data processing regions and prepares their data in advance, ensuring that data is ready and available when needed. This preliminary action prevents processing stops and maintains productivity while avoiding memory fragmentation through proactive memory management.
Solution Approach 2:
The buffer memory management system ensures continuous data processing by maintaining multiple regions with different data states. The first region provides continuous data for current processing, the second region provides continuous data for next regions, and the third region provides continuous workspace. This continuity of useful action prevents processing stops and maintains reliability while supporting high productivity through uninterrupted data flow.
3Productivity
If data is stored in buffer memory for multiple data processing regions, then productivity is improved, but memory fragmentation occurs leading to swap-out operations
Solution Approach 1:
The buffer memory is segmented into distinct regions with specific functions: the first region for current processing data, the second region for next region data, and the third region for work area. This segmentation simplifies memory management by providing clear boundaries and purposes for each region, preventing fragmentation while maintaining high data processing speed. The structured approach reduces management complexity compared to unsegmented memory.
Solution Approach 2:
The system dynamically changes memory parameters by adjusting the allocation and usage of different buffer memory regions based on processing needs. The control unit manages data movement between regions and adjusts working area sizes as needed. This dynamic parameter change enables efficient memory utilization and fast data processing while avoiding fragmentation through adaptive memory management.
Data Source
AI summary
A charged particle beam writing apparatus includes a buffer memory including a memory region capable of contemporarily storing writing data for data processing regions, wherein writing data including data files is temporarily stored for each of the data processing regions, a dividing unit to divide the memory region of the buffer memory into a first region being large and a second region being small, a specifying unit to specify the memory region such that a data file being large is preferentially stored in the first region and a data file being small is stored at least in the second region, concerning the data files for each of the data processing regions included in the writing data, and a data processing unit to read data files corresponding to each of the data processing regions from the buffer memory, and to perform data processing using the read data files.