Lithographic Apparatus Buffer System for Data Rate Bottleneck
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional lithographic apparatuses face challenges in rapidly reconfiguring patterning devices for maskless lithography due to difficulties in programming pattern data at high data rates, making the programming of patterning devices a rate-limiting step in the lithographic process.
Innovation Solution
The method involves dividing the pattern data into regions, assigning unique identifiers to each region, and storing only one instance of the pattern per identifier, with a buffer system for efficient data transfer to the control device, utilizing a resource schedule for optimized data processing and memory management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the patterning device is reconfigured rapidly to write different pixels to different target portions, then the productivity of the lithographic process is improved, but the programming of the patterning device becomes a rate-limiting step due to difficulty in handling large data volumes
Solution Approach 1:
The pattern data is divided into multiple tiles, where each tile represents a portion of the overall pattern. This segmentation allows the system to process and transfer smaller data units independently, reducing the complexity of handling the complete pattern data at once while enabling rapid reconfiguration of the patterning device for different target portions.
2Manufacturing precision
If the complete pattern data is stored and transferred to the patterning device, then the manufacturing precision is maintained, but the memory requirements and data transfer volume become excessively large
Solution Approach 1:
Tiles that are needed for upcoming writing operations are loaded into memory in advance, before they are actually needed by the patterning device. This preliminary action allows the system to maintain high manufacturing precision by having the required pattern data ready, while avoiding the need to store and transfer the complete pattern data set, thus reducing memory requirements.
3Productivity
If all pattern data is loaded into memory, then the data processing efficiency is improved, but the system costs and memory requirements increase substantially
Solution Approach 1:
Instead of uniformly loading all pattern data into memory, the system selectively loads only those tiles that are currently needed or will be needed soon for specific target portions. This local quality approach ensures that memory resources are concentrated on the relevant data subsets, maintaining high data processing efficiency for active operations while substantially reducing overall memory storage capacity requirements.
Data Source
AI summary
A method and device for programming an array of individually controllable elements configured to impart a beam with a pattern. For example, the method can be suitable for use in a lithographic apparatus. The method includes generating first data representing a first pattern, generating second data representing a second pattern, writing the first data to a first buffer, and reading the first data from the first buffer to program the array of individually controllable elements to display the first pattern, while writing the second data to a second buffer in parallel.