MEBESZIP
MEBESZIP addresses the challenge of handling increasingly large MEBES files sizes in the nanometer design era. It converts mask data in MEBES format into a highly compressed proprietary binary format. The binary format can be decompressed back to recover the original bit-for-bit matching MEBES file.Why MEBESZIP-Large Mask Data Files Need Compression
MEBES (Manufacturing Electron Beam Exposure System) is a proprietary mask data format from Applied Materials Inc. for exchange of fractured mask data. It is the most widely used format for representing and exchanging mask data and is considered as a de facto standard format in photomask industry. As per the ITRS roadmap of 2005, for a 45nm half-pitch node that is expected to be in use by 2010, the mask data volume for a single layer is expected to reach up to 825 GB. The industry needs solutions that can deal with this enormous data explosion. Compression techniques better than generic and commonly used compressors like gzip will be required to address this problem. In recent benchmarks, comparing against generic compressors such as gzip, the MEBESZIP has been consistently producing compression factors ranging between 2X and 8X better than gzip, with an average of 4X better compression across all tests runs. Compression ratios produced by MEBESZIP are significantly better than any other similar products.
When compared to post-layout data formats such as GDSII and OASIS, mask data is much more voluminous and more appropriate candidate for data compression due to various factors such as:
- The input data to the e-beam machines requires the fractured mask data to be essentially flat (very limited amount of hierarchy) which will bloat the size of the polygonal data.
- Techniques such as RET (Resolution enhancement Techniques), or dummy fills, which are commonly required in current designs to improve yield, affect the volume of mask data more than layout data.
- By its very nature, the process of Fracturing, i.e. the operation that generates mask data from layout data creates several geometries out of what was originally a single polygon.
- Unlike GDSII, MEBES already has built into the language both syntactic and semantic representations that are more compact for representing fractured mask data. Because of this, generic compressors such as gzip are not able to produce as much compression on mask data as they are able to produce on layout data such as GDSII.
Key features and benefits of MEBESZIP
- Cost and time savings: By compressing MEBES files by a factor of 5-15X, MEBESZIP reduces the time taken to transfer these files across mask shops, foundries and manufacturing-related groups within Integrated Device Manufacturers (IDMs). MEBESZIP reduces the disk space requirements for storing and archiving of MEBES files within these organizations.
- Integration with post-layout EDA Tools: MEBESZIP can be integrated as a technology component of EDA tools, such as Mask Data Preparation (MDP), Resolution Enhancement Technology (RET), Manufacturing Rules Check (MRC), Mask Data Viewer and Mask Inspection tools that generate or consume MEBES data by using the Application Programming Interface (API) available with MEBESZIP. It speeds up the entire mask data preparation, analysis and verification flow by eliminating the need for reading and writing entire MEBES files.
- Selective Compression and De-Compression: MEBESZIP allows for segment-wise selective compression and decompression of MEBES files. This feature in conjunction with API-based integration with upstream and downstream EDA tools enables on-demand reading of compressed MEBES data without having to decompress the entire file before being able to operate on the data.
- Verification and Security: MEBESZIP provides an optional in-built verifier that provides on-the-spot verification to check the fidelity of the compressed file. CRC-32 based checksum and MD-5 based fingerprint schemes are used in MEBESZIP for verifying the data integrity. MEBESZIP offers the option of in-built data encryption and decryption to enhance the security during transfer of compressed data.