Memory State Compressors for Giga-Scale Checkpoint/Restore

We propose a checkpoint store compression method for coarse-grain giga-scale checkpoint/restore. This mechanism can be useful for debugging, post-mortem analysis and error recovery. Our compression method exploits value locality in the memory data and address streams. Our compressors require few resources, can be easily pipelined and can process a full cache block per processor cycle. We study two applications of our compressors for post-mortem analysis: (1) Using them alone, and (2) using them in-series with a dictionary-based compressor. When used alone they offer competitive compression rates in most cases. When combined with dictionary compressors, they significantly reducing onchip buffer requirements.

• Type: Conference
• Year: 2005
• Where: IEEEPACT
• Authors: Andreas Moshovos, Alexandros Kostopoulos
• Keywords: Compression Method, Hardware, IEEEPACT 2005, Post-mortem Analysis, Store Compression Method