Operation method of storage device configured to support multi-stream
Disclosed is an operation method of a storage device supporting a multi-stream, which includes receiving an input/output request from an external host, generating a plurality of stream identifier candidates by performing machine learning on the input/output request based on a plurality of machine learning models that are based on different machine learning algorithms, generating a model ratio based on a characteristic of the input/output request, applying the model ratio to the plurality of stream identifier candidates to allocate a final stream identifier for the input/output request, and storing write data corresponding to the input/output request in a nonvolatile memory device of the storage device based on the final stream identifier.
1. An operation method of a storage device which supports a multi-stream, the method comprising:
receiving, by a storage controller circuit, an input/output request from an external host;
generating, by the storage controller circuit, a plurality of stream identifier candidates by performing machine learning on the input/output request based on a plurality of machine learning models that are based on different machine learning algorithms;
generating, by the storage controller circuit, a model ratio based on a characteristic of the input/output request;
applying, by the storage controller circuit, the model ratio to the plurality of stream identifier candidates to allocate a final stream identifier for the input/output request; and
storing, by the storage controller circuit, write data corresponding to the input/output request in a nonvolatile memory device of the storage device based on the final stream identifier.
2. The method as claimed in claim 1 ,
wherein the characteristic of the input/output request includes at least one of a logical block address, a logical block address range, and a request size of the input/output request.
3. The method as claimed in claim 1 ,
wherein the characteristic of the input/output request includes at least one of an overwrite ratio and a logical block address continuity of input/output requests adjacent to the input/output request.
4. The method as claimed in claim 1 ,
wherein the model ratio includes a plurality of coefficients respectively corresponding to the plurality of machine learning models, and
wherein the plurality of coefficients have different values.
5. The method as claimed in claim 4 ,
wherein the plurality of coefficients when the characteristic of the input/output request corresponds to a first logical block address range are different from the plurality of coefficients when the characteristic of the input/output request corresponds to a second logical block address range.
6. The method as claimed in claim 1 ,
wherein each of the plurality of machine learning models is based on at least one of a linear regression algorithm, a support vector machine algorithm, a deep neural network algorithm, a deep stream algorithm, a K-Mean algorithm, a clustering algorithm, an auto encoder algorithm, a convolutional neural network algorithm, and a Siamese network algorithm.
7. The method as claimed in claim 1 ,
wherein the generating of the model ratio based on the characteristic of the input/output request includes:
extracting the characteristic of the input/output request; and
selecting and outputting the model ratio corresponding to the characteristic from a lookup table, and
wherein the lookup table includes a plurality of entries, and each of the plurality of entries includes information about a model ratio corresponding to each of a plurality of characteristics.
8. The method as claimed in claim 7 , wherein the lookup table is included in firmware of the storage device.
9. The method as claimed in claim 1 ,
wherein, when the final stream identifier is a first stream identifier, the write data are stored in a first memory block of the nonvolatile memory device, and
wherein, when the final stream identifier is a second stream identifier different from the first stream identifier, the write data are stored in a second memory block of the nonvolatile memory device, which is different from the first memory block.
10. The method as claimed in claim 1 ,
wherein the input/output request includes accumulated write requests, and a size of accumulated write data corresponding to the accumulated write requests is a given size.
11. The method as claimed in claim 1 , wherein the input/output request includes accumulated write requests, and a number of the accumulated write requests is a given number.