Presentation
ASMCap: An Approximate String Matching Accelerator for Genome Sequence Analysis Based on Capacitive Content Addressable Memory
DescriptionGenome sequence analysis is a foundation in medical and scientific areas. Considering the inevitable errors, Approximate String Matching (ASM) is more practical due to its ability to tolerate errors. However, ASM also heavily degrades the energy efficiency of existing accelerators.
In this paper, we present a novel ASM accelerator, namely ASMCap, based on the capacitive Multi-Level Content Addressable Memories (ML-CAMs). The specific matching pattern of ASMCap can tolerate the edit distance, and the charge-domain computation mode can achieve high reliability and reduce power consumption with no area penalty. In addition, we also propose two strategies and analyze their overheads.
In this paper, we present a novel ASM accelerator, namely ASMCap, based on the capacitive Multi-Level Content Addressable Memories (ML-CAMs). The specific matching pattern of ASMCap can tolerate the edit distance, and the charge-domain computation mode can achieve high reliability and reduce power consumption with no area penalty. In addition, we also propose two strategies and analyze their overheads.
Event Type
Research Manuscript
TimeTuesday, July 11th3:40pm - 3:55pm PDT
Location3010, 3rd Floor
Design
In-memory and Near-memory Computing Circuits