Data storage device and method for reducing flush latency
A data storage device has a cache and a non-volatile memory. Instead of flushing the entire cache to the non-volatile memory in response to a command from a host, the data storage device flushes only the cached data that is associated with an identifier provided by the host. This allows the cached data associated with the identifier to be flushed more quickly. The data storage device can also prioritize queued commands that are associated with the identifier.
1. A data storage device comprising:
a non-volatile memory; and
a processor coupled with the non-volatile memory and configured to:
store, in a cache, data from a plurality of applications in a host to be written in the non-volatile memory, wherein each data is associated with a respective identifier;
receive, from the host, a flush command comprising a first identifier; and
in response to receiving the flush command comprising the first identifier:
select, from a command queue, only write command(s) that are associated with the first identifier;
store data associated with the selected write command(s) in the cache; and
flush, from the cache to the non-volatile memory, only data that is associated with the first identifier, wherein the data that is flushed to the non-volatile memory comprises data stored in the cache both before and after the flush command was received.
2. The data storage device of claim 1 , wherein the first identifier comprises a context identifier, a stream identifier, a zone identifier, a group number identifier, and/or an application identifier.
3. The data storage device of claim 1 , wherein the processor is further configured to store identifiers in a table, in a linked list, and/or with the data in the cache.
4. The data storage device of claim 1 , wherein the processor is further configured to store data associated with different identifiers in different address ranges in a volatile memory.
5. The data storage device of claim 1 , wherein the processor is further configured to store data associated with different identifiers in different single level cell (SLC) blocks in the non-volatile memory.
6. The data storage device of claim 1 , wherein a cache size is larger than a page size of the non-volatile memory and flushing the cache comprises performing a plurality of write operations.
7. The data storage device of claim 1 , wherein the cache is part of a volatile memory of the data storage device.
8. The data storage device of claim 1 , wherein the processor is further configured to arrange the data in the cache according to temperature of the data, zone, stream, or identifier.
9. The data storage device of claim 1 , wherein the non-volatile memory comprises a three-dimensional memory.
10. In a data storage device comprising non-volatile memory, a method comprising:
caching, in a cache, data received from a plurality of applications in a host, wherein each data is associated with a respective identifier that identifies the application from which the data originated;
receiving, from the host, a synch cache command comprising an identifier of one of the plurality of applications, wherein the host sends the synch cache command to the data storage device to confirm that data from the one of the plurality of applications has been stored in the non-volatile memory; and
in response to receiving the synch cache command:
using the identifier to identifying which of the data stored in the cache is associated with the one of the plurality of applications;
copying only the data associated with the one of the plurality of applications from the cache to the non-volatile memory, wherein the data that is copied to the non-volatile memory comprises any data associated with the one of the plurality of applications that was stored in the cache both before and after the synch cache command was received; and
informing the host that the data associated with the one of the plurality of applications has been stored in the non-volatile memory, wherein copying only the data associated with the one of the plurality of applications from the cache to the non-volatile memory reduces latency in informing the host that the data associated with the one of the plurality of applications has been stored in the non-volatile memory.
11. The method of claim 10 , further comprising:
prioritizing a command in a command queue of the data storage device that is associated with the identifier.
12. The method of claim 10 , wherein the identifier comprises a context identifier, a stream identifier, a zone identifier, a group number identifier, and/or an application identifier.
13. The method of claim 10 , wherein data cached in the cache is associated with different identifiers, and wherein the different identifiers are stored in a table, in a linked list, and/or with the data in a cache.
14. The method of claim 10 , wherein data cached in the cache is associated with different identifiers, and wherein the different identifiers are stored in different address ranges in a volatile memory.
15. The method of claim 10 , wherein data cached in the cache is associated with different identifiers, and wherein the different identifiers are stored in different single level cell (SLC) blocks in the non-volatile memory.
16. The method of claim 10 , wherein a cache size is larger than a page size of the non-volatile memory and copying comprises performing a plurality of write operations.
17. The method of claim 10 , wherein the cache is in a volatile memory of the data storage device and/or in the non-volatile memory.
18. The method of claim 10 , further comprising arranging the data in the cache according to temperature of the data, zone, stream, or identifier.
19. A data storage device comprising:
a cache;
a non-volatile memory; and
means for:
storing, in the cache, data from a plurality of applications in a host to be written in the non-volatile memory, wherein each data is associated with a respective identifier;
receiving, from the host, a flush command comprising a first identifier; and
in response to receiving the flush command comprising the first identifier:
selecting, from a command queue, only write command(s) that are associated with the first identifier;
storing data associated with the selected write command(s) in the cache; and
flushing, from the cache to the non-volatile memory, only data that is associated with the first identifier, wherein the data that is flushed to the non-volatile memory comprises data stored in the cache both before and after the flush command was received.
20. The data storage device of claim 1 , wherein the cache is part of the non-volatile memory of the data storage device.