GPU-Based Redundancy Analysis using Partitioning Method for Memory Repair | IEEE Conference Publication | IEEE Xplore

GPU-Based Redundancy Analysis using Partitioning Method for Memory Repair


Abstract:

Redundancy Analysis (RA) is an effective memory repair method to improve memory yield. However, conventional RA methods with simple spare structures have limitations in a...Show More

Abstract:

Redundancy Analysis (RA) is an effective memory repair method to improve memory yield. However, conventional RA methods with simple spare structures have limitations in achieving optimal repair rates since local spares can only repair the memory banks they belong to. To overcome this penalty, research on complex spare structures has been conducted. However, generating solutions for complex spare structures can be time-consuming when dealing with multiple memories. Therefore, a new complex spare structure is proposed using GPU parallel processing and redundancy group partitioning. Experimental results show that the proposed method achieves a higher repair rate with a similar RA time compared to conventional GPU-based RA with a simple spare structure.
Date of Conference: 25-28 October 2023
Date Added to IEEE Xplore: 24 January 2024
ISBN Information:

ISSN Information:

Conference Location: Jeju, Korea, Republic of

References

References is not available for this document.