Storage system and method for circuit-bounded-array-based time and temperature tag management and inference of read thresholds
A storage system has an inference engine that can infer a read threshold based on a plurality of parameters of the memory. The read threshold can be used in reading a wordline in the memory during a regular read operation or as part of an error handling process. Using a machine-learning-based approach to infer a read threshold can provide significant improvement in read threshold accuracy, which can reduce bit error rate and improve latency, throughput, power consumption, and quality of service. In another embodiment, a circuit-bounded array is used to manage updates to time and temperature tag information and to infer read thresholds.
1. A storage system comprising:
a memory comprising a plurality of memory dies, wherein each memory die comprises a respective circuit-bounded array; and
a processor coupled with the memory and configured to perform the following for each of the plurality of memory dies:
read a wordline in a memory die;
determine a read threshold based on the read wordline; and
send the read threshold to the circuit-bounded array in the memory die;
wherein each circuit-bounded array is configured to apply a machine-learning-based adjustment to the read threshold.
2. The storage system of claim 1 , wherein the wordline read by the processor is part of a time and temperature group and the circuit-bounded array is further configured to apply the machine-learning-based adjustment to the read threshold in response to the time and temperature group being read.
3. The storage system of claim 2 , wherein the processor is further configured to manage time and temperature groups of all of the plurality of memory dies.
4. The storage system of claim 3 , wherein the processor is further configured to determine the read thresholds in response to a time and temperature group update.
5. The storage system of claim 1 , wherein each circuit-bounded array is further configured to apply the machine-learning-based adjustment to the read threshold on-the-fly before reading a target wordline.
6. The storage system of claim 1 , wherein at least one circuit-bounded array uses a different machine-learning algorithm than another one of the circuit-bounded arrays.
7. The storage system of claim 1 , wherein at least one circuit-bounded array uses a machine-learning algorithm that is trained at least in part off-line.
8. The storage system of claim 1 , wherein each circuit-bounded array comprises a respective memory chip and a separate respective complementary metal oxide semiconductor (CMOS) chip.
9. The storage system of claim 1 , wherein at least one of the circuit-bounded arrays is a Circuit Above the Array (CAA) device.
10. The storage system of claim 1 , wherein at least one of the circuit-bounded arrays is a Circuit Under the Array (CUA) device.
11. The storage system of claim 1 , wherein the memory comprises a three-dimensional memory.
12. In a storage system comprising a memory comprising a plurality of circuit-bounded array (CBAs), wherein each CBA includes a memory die, a method comprising:
performing the following in each CBA:
reading a wordline in a time and temperature group in the memory die of the CBA;
determining a read threshold based on the read wordline;
storing the read threshold in the CBA; and
applying a machine-learning-based adjustment to the read threshold.
13. The method of claim 12 , wherein each CBA applies the machine-learning- based adjustment to the read threshold on-the-fly before reading a target wordline.
14. The method of claim 12 , wherein each CBA comprises the respective memory die and a separate respective complementary metal oxide semiconductor (CMOS) chip coupled to the memory die.
15. The method of claim 12 , wherein each CBA determines the read threshold based on the read wordline in parallel with at least one other CBA.
16. The method of claim 12 , wherein at least one of the CBAs reads the wordline in the time and temperature group during idle time of its memory die.
17. The method of claim 12 , wherein at least one of the CBAs reads the wordline in the time and temperature group in response to receiving a command from a storage controller of the storage system.
18. The method of claim 12 , wherein at least one CBA uses a different machine-learning algorithm than another one of the CBAs.
19. The method of claim 12 , wherein at least one CBA uses a machine-learning algorithm that is trained at least in part off-line.
20. A storage system comprising:
a memory comprising a circuit-bounded array (CBA) including a memory die; and
means, located in the circuit-bounded array, for:
reading a wordline in a time and temperature group in the memory die of the CBA;
determining a read threshold based on the read wordline;
storing the read threshold in the CBA; and
applying a machine-learning-based adjustment to the read threshold.