Method of operating a storage device and a memory controller
By using a reinforcement learning model to optimize garbage collection strategies in storage devices and selecting appropriate sacrifice blocks and target blocks, the problems of low efficiency and shortened lifespan of storage devices during data writing are solved, achieving more efficient garbage collection and a longer device lifespan.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SAMSUNG ELECTRONICS CO LTD
- Filing Date
- 2021-02-01
- Publication Date
- 2026-07-07
AI Technical Summary
Existing storage devices are inefficient in data writing and have poor garbage collection efficiency, which leads to a shortened lifespan of the storage devices.
A reinforcement learning model is adopted to select appropriate sacrificial blocks and target blocks for garbage collection based on the state information and data feature information of storage blocks. The strategy is optimized through the FTL layer of the memory controller.
It improves the efficiency of garbage collection, reduces the write amplification factor of storage devices, extends the lifespan of storage devices, and enhances write operation performance.
Smart Images

Figure CN113254362B_ABST