content-aware-ftl · research · also CAFTL, Content-aware flash translation layer

Content-aware FTL

A content-aware flash translation layer uses data fingerprints to share duplicate pages and suppress redundant writes inside an SSD.

In the CAFTL research design, inline deduplication can avoid programming a duplicate. Background deduplication merges copies already stored; it cannot undo their earlier writes. A bounded fingerprint store and sampling can miss duplicates, so this is best-effort reduction. [1]

Shared pages use logical-to-virtual-to-physical mapping. Relocation updates the shared virtual mapping, while an exact reference count prevents reclamation until no logical page refers to the physical page. The fingerprint cache's popularity counter is a separate quantity, not the reclamation authority. [1]

CRC32 pre-hashing only screens candidates before the design's SHA-1 fingerprint comparison; CRC equality alone does not authorize sharing. The historical SHA-1 assumption is not a current security recommendation. [1] [2]

Relations

related

Flash translation layer (FTL), Data deduplication, Write amplification, Hash collision, Garbage collection (flash), SSD mapping cache

Limitations

• Deduplication savings do not by themselves give total device write amplification or prove power-loss safety. [1]

Citations

1. Chen, Luo and Zhang. CAFTL: A Content-Aware Flash Translation Layer Enhancing the Lifespan of Flash Memory based Solid State Drives. FAST 2011, sections 3-4. (accessed 2026-11-08)
2. NIST CSRC. Hash Functions: SHA-1 deprecation and collision resistance. (accessed 2026-11-08)

JSON: /data/terms/content-aware-ftl.json