Refresh and access modes for memory
View Patent ↗The present disclosure includes apparatuses and methods related to copying data in a memory system with an artificial intelligence (AI) mode. An apparatus can receive a command indicating that the apparatus operate in an artificial intelligence (AI) mode, a command to perform AI operations using an AI accelerator based on a status of a number of registers, and a command to copy data between memory devices that are performing AI operations. The memory system can copy neural network data, activation function data, bias data, input data, and/or output data from a first memory device to a second memory device, such that that the first memory device can use the neural network data, activation function data, bias data, input data, and/or output data in a first AI operation and the second memory device can use the neural network data, activation function data, bias data, input data, and/or output data in a second AI operation.
1. An apparatus, comprising:
a number of memory devices; and
a controller coupled to the number of memory devices and configured to:
receive a command comprising a first address identifying a first memory device and a second address identifying a second memory device of the number of memory devices as a target for data stored in the first memory device;
transmit data from the first memory device to the second memory device in response to the command; and
execute a training operation, an inference operation, or both, at the second memory device using the stored data.
2. The apparatus of claim 1 , wherein each of the number of memory devices is configured as part of a neural network and includes a number of memory arrays.
3. The apparatus of claim 1 , further comprising the controller configured to provide input information associated with data stored in the first memory device for the training operation.
4. The apparatus of claim 1 , further comprising the controller configured to provide input information for the inference operation.
5. The apparatus of claim 1 , further comprising the controller configured to send a command to copy data from the first memory device to the second memory device,
wherein the first memory device and the second memory device perform artificial intelligence (AI) operations using the data.
6. The apparatus of claim 5 , wherein the copied data is neural network data.
7. The apparatus of claim 1 , wherein the command instructs the second memory device to receive the stored data and write the stored data to a memory array in the second memory array.
8. A system, comprising:
a number of memory devices; and
a controller coupled to the number of memory devices and configured to:
receive a first command to execute a first training or inference operation on a first memory device of the number of memory devices;
receive a second command to:
copy data from the first memory device to a second memory device of the number of memory device; and
execute a second training or inference operation on the second memory device, wherein the second command comprises a first address and a second address, the first address identifying the first memory device storing data associated with the first training or inference operation and the second address identifying the second memory device as a target for the data.
9. The system of claim 8 , comprising the controller configured to transmit data from the first memory device to the second memory device in response to the first command, the second command, and input information comprising a set of data providing information associated with the copied data.
10. The system of claim 8 , comprising the controller configured to execute the training or inference operation at the second memory device using the copied data.
11. The system of claim 8 , further comprising a host device, wherein the host device is configured to issue the command.
12. The system of claim 8 , further comprising the second memory device to store input information associated with data stored in a first memory device for the first training operation, the inference operation, or both.
13. The system of claim 11 , further comprising the host to store input information associated with data stored in a first memory device for the first training operation, the inference operation, or both.
14. A method, comprising:
receiving, from a host device and at a first memory device of a plurality of memory devices, a command comprising a first address and a second address, the first address identifying the first memory device storing data for a training operation, an inference operation, or both, and the second address identifying a second memory device of the plurality of memory devices as a target for the stored data; and
transmitting the stored data from the first memory device to the second memory device in response to the command.
15. The method of claim 14 , comprising receiving, at the first memory device input information associated with the stored data for the training operation, the inference operation, or both, wherein the stored data comprises the input information for the training operation, the inference operation, or both.
16. The method of claim 14 , comprising executing, at the second memory device, the training operation, the inference operation, or both.
17. The method of claim 15 , comprising the second memory device storing the input information.
18. The method of claim 15 , further comprising transmitting the stored data from the first memory device to the second memory device based on the input information.
19. The method of claim 15 , wherein transmitting the stored data from the first memory device to the second memory device comprises copying the input information for the training operation, the inference operation, or both.
20. The method of claim 14 , wherein transmitting the stored data from the first memory device to the second memory device comprises copying neural network data for the training operation, the inference operation, or both.