Apparatus and method for adjusting cache allocated for read look ahead

A controller, comprising:

• at least one register configured to store a doorbell regarding a submission queue for at least one request generated by a host;
• a read cache including a default area for an operation performed in response to the at least one request and a variable area for an operation performed in response to a read look ahead (RLA) request generated based on the at least one request;
• and
• a cache size manager configured to determine a cache size for the RLA request based on an update cycle of the doorbell and adjust a size of the variable area allocated for the RLA request based on the determined cache size.

A controller, comprising:

• a plurality of registers configured to store a submission queue head doorbell and a submission queue tail doorbell regarding a plurality of requests generated by a host;
• a read cache including a first area and a second area, wherein the first area is configured to store data corresponding to the plurality of requests, and at least a part of the second area is allocated to store data corresponding to a read look ahead (RLA) request generated based on the plurality of requests;
• and
• a cache size manager configured to determine a cache size for the RLA request based on a difference of the submission queue head doorbell and the submission queue tail doorbell and adjust a size of the at least part of the second area allocated for the RLA request based on the determined cache size.

A method for operating a memory system, comprising:

• storing a doorbell regarding a submission queue storing at least one request generated by a host;
• storing first data read from a non-volatile memory corresponding to the at least one request in a default area of a read cache;
• generating a read look ahead (RLA) request based on the at least one request;
• storing second data read from the non-volatile memory corresponding to the RLA request in a variable area allocated for the RLA request in the read cache;
• determining a cache size for the RLA request based on an update cycle of the doorbell;
• and
• adjusting a size of the variable area based on the determined cache size for changing an operation throughput of the RLA request.

A memory system, comprising:

• a non-volatile memory for storing data;
• at least one register configured to store a doorbell regarding a submission queue for at least one request generated by a host;
• a read cache including a first area and a second area, wherein the first area is allocated to store first data, read from the non-volatile memory corresponding to the at least one request, and at least some part of the second area is allocated to store second data read from the non-volatile memory corresponding to a read look ahead (RLA) request generated based on the at least one request;
• and
• a cache size manager configured to determine a cache size for the RLA request based on a read performance of the memory system and adjust a size of the at least some part of the second area allocated for the RLA request based on the determined cache size.