Charged Particle Beam Writing Pipeline Data Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing data volume in electron beam lithography for semiconductor manufacturing leads to prolonged data generation and transmission times, resulting in increased pattern writing times and manufacturing costs, as existing technologies rely on external devices for data conversion and processing.
Innovation Solution
A charged particle beam writing apparatus with multiple storage devices and data processor units that perform pipeline processing to convert design data into draw data efficiently, reducing the need for external data conversion and enabling simultaneous data processing and transfer within the apparatus.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If data conversion and processing are performed by external devices before transferring to the electron beam pattern writing apparatus, then the apparatus can receive pre-processed draw data, but the total time for data generation and transmission is significantly increased
Solution Approach 1:
The patent merges the data conversion and processing functions from external devices into the electron beam pattern writing apparatus itself. The draw data creation unit and multiple data processing units are integrated within the apparatus, allowing data conversion and processing to occur simultaneously with data transfer, thereby eliminating the sequential time loss of external processing followed by transfer.
Solution Approach 2:
The apparatus performs data conversion and processing operations in advance during the data transfer process itself. By implementing pipeline processing where data is converted and processed while being transferred, the system prepares data for pattern writing without adding separate time steps, thus reducing overall pattern writing time.
2Reliability
If all draw data are transferred in unison to the electron beam pattern writing apparatus, then the apparatus can process complete datasets, but the transmission time is appreciably increased
Solution Approach 1:
The patent segments the data transfer process into multiple parallel streams using a pipeline configuration. Multiple data processing units (first through n-th units) process different portions of data simultaneously, allowing the complete dataset to be transferred and processed in divided segments rather than waiting for all data to arrive before processing begins.
Solution Approach 2:
The pipeline processing enables continuous data transfer and processing operations. While data is being transferred, multiple processing units continuously work on different segments of the data simultaneously, eliminating idle time and ensuring that the useful action of data processing continues throughout the entire transfer period rather than occurring only after complete transfer.
3Productivity
If the electron beam pattern writing apparatus processes huge amounts of data, then higher integration density can be achieved, but the manufacturing costs of masks increase
Solution Approach 1:
The electron beam pattern writing apparatus performs its own data conversion and processing operations through integrated data processing units, eliminating the need for external device support. This self-service capability reduces operational complexity and associated costs while maintaining the ability to process huge datasets required for high integration density patterns.
Data Source
AI summary
A charged-particle beam writing apparatus includes first and second storage devices, a transfer processor for sequentially sending to the first storage several design data files per pattern layout-defined region, a first data processor which sequentially reads design data files from the first storage and converts each data file's design data into draw data for storage in the second storage while being pipelined with the transfer processing, second to n-th data processors which sequentially read data files from the second storage and apply mutually different ones of second to n-th data processings to each draw data in a way that the first to n-th data processings are pipelined and store the processed draw data in the second storage, and a pattern-writing unit for writing a pattern on a workpiece by using a beam that is controlled based on each n-th data processing-completed data being stored in the second storage.