Flash Controller H2F Sub-Table Update for Random Write

The patent segments the administration table programming task by dividing it into multiple sub-tables (H2F sub-tables) that can be programmed in parallel across multiple channels. Instead of programming the entire administration table sequentially through a single channel, the controller divides the work into manageable segments that can be processed simultaneously by different channels, thereby improving programming speed while maintaining the simplicity of serial access within each channel.

The patent applies preliminary action by determining whether user data satisfies random-write conditions before initiating the administration table programming process. The controller evaluates the write pattern characteristics in advance and only updates H2F sub-tables when necessary (i.e., when random-write conditions are detected). This prevents unnecessary programming operations and optimizes system performance by performing administration table updates only when they will actually improve subsequent read/write operations.

The patent implements periodic action by updating H2F sub-tables at specific intervals or under specific conditions rather than continuously or at fixed time points. The controller periodically checks whether user data satisfies random-write conditions and only updates the administration tables when such conditions are met. This conditional periodic updating ensures data access accuracy is maintained while minimizing unnecessary update operations that would cause delays and timeouts.

The patent applies self-service by enabling the flash controller to autonomously determine when administration table updates are necessary based on the characteristics of incoming write operations. The controller monitors write patterns itself and automatically decides whether to update H2F sub-tables based on whether random-write conditions are detected, without requiring external intervention or timing signals. This self-service approach optimizes the balance between data access accuracy and update timing.

The patent segments the administration table management task by dividing it into multiple H2F sub-tables, each associated with specific channels or data ranges. This segmentation allows each channel to manage its own relevant sub-tables independently, reducing the complexity of coordinating updates across all channels while maintaining parallel processing capabilities. Each channel can update its associated sub-tables without requiring complex inter-channel coordination.

The patent applies local quality by making the administration table management structure adaptive to local channel characteristics and data patterns. Instead of using a uniform management approach for all channels, the controller evaluates random-write conditions locally for each channel's data and only updates sub-tables where necessary. This localized management approach reduces overall complexity by focusing updates only where they are needed while maintaining high throughput through parallel channel operations.