System and method for discovering parallelism of memory devices
A data processing system includes a host device and a memory system including a plurality of units. The host device includes a workload generation component and an analysis component. The workload generation component concurrently transmits, to the memory system, a plurality of commands for the plurality of memory units. The analysis component receives, from the memory system, command completion messages corresponding to the plurality of commands; measures latencies of the plurality of commands based on the receiving of the command completion messages; and analyze a parallelism scheme of the plurality of memory units based the measured latencies.
1. A data processing system comprising:
a host device; and
a memory system including a plurality of memory units,
wherein the host device includes:
a workload generation component configured to concurrently transmit, to the memory system, a plurality of commands for the plurality of memory units; and
an analysis component configured to receive, from the memory system, command completion messages corresponding to the plurality of commands; measuring latencies of the plurality of commands based on the receiving of the command completion messages; and analyzing a parallelism scheme of the plurality of memory units based the measured latencies.
2. The data processing system of claim 1 , wherein the plurality of commands includes read commands.
3. The data processing system of claim 2 , wherein the workload generation component further transmits sequential write commands to the plurality of memory units before transmitting the read commands.
4. The data processing system of claim 3 , wherein the workload generation component further transmits erase commands to the plurality of memory units before transmitting the write commands.
5. The data processing system of claim 1 , wherein the analysis component determines normalized delta latencies between two latencies among the measured latencies, and identifies the parallelism scheme of the plurality of memory units based on patterns of the normalized delta latencies.
6. The data processing system of claim 5 , wherein patterns of the normalized delta latencies are identical to each other for the same memory unit among the plurality of memory units.
7. The data processing system of claim 6 , wherein patterns of the normalized delta latencies are different for the different memory units among the plurality of memory units.
8. The data processing system of claim 1 , wherein the parallelism scheme includes the number of the plurality of memory units and a striping unit of each of the plurality of memory units.
9. The data processing system of claim 1 , wherein the plurality of memory units includes a plurality of memory dies.
10. A method for operating a host device coupled to a memory system including a plurality of memory units, the method comprising:
concurrently transmitting, to the memory system, a plurality of commands for the plurality of memory units;
receiving, from the memory system, command completion messages corresponding to the plurality of commands;
measuring latencies of the plurality of commands based on the receiving of the command completion messages; and
analyzing a parallelism scheme of the plurality of memory units based the measured latencies.
11. The method of claim 10 , wherein the plurality of commands includes read commands.
12. The method of claim 11 , further comprising:
transmitting sequential write commands to the plurality of memory units before transmitting the read commands.
13. The method of claim 12 , further comprising:
transmitting erase commands to the plurality of memory units before transmitting the write commands.
14. The method of claim 10 , wherein the analyzing of the parallelism scheme of the plurality of memory units comprises:
determining normalized delta latencies between two latencies among the measured latencies; and
identifying the parallelism scheme of the plurality of memory units based on patterns of the normalized delta latencies.
15. The method of claim 14 , wherein patterns of the normalized delta latencies are identical to each other for the same memory unit among the plurality of memory units.
16. The method of claim 15 , wherein patterns of the normalized delta latencies are different for the different memory units among the plurality of memory units.
17. The method of claim 10 , wherein the parallelism scheme includes the number of the plurality of memory units and a striping unit of each of the plurality of memory units.
18. The method of claim 10 , wherein the plurality of memory units includes a plurality of memory dies.