Systolic array-friendly data placement and control based on masked write
The present disclosure relates to an accelerator for systolic array-friendly data placement. The accelerator may include: a systolic array comprising a plurality of operation units, wherein the systolic array is configured to receive staged input data and perform operations using the staged input to generate staged output data, the staged output data comprising a number of segments; a controller configured to execute one or more instructions to generate a pattern generation signal; a data mask generator; and a memory configured to store the staged output data using the generated masks. The data mask generator may include circuitry configured to: receive the pattern generation signal from the controller, and, based on the received signal, generate a mask corresponding to each segment of the staged output data.
1. An accelerator comprising:
a systolic array comprising a plurality of operation units, wherein the systolic array is configured to receive staged input data and perform operations using the staged input to generate staged output data, the staged output data comprising a number of segments;
a controller configured to execute one or more instructions to generate a pattern generation signal based on a type of data output by the systolic array;
a data mask generator including circuitry configured to:
receive the pattern generation signal from the controller, and
based on the received signal, generate a mask corresponding to each segment of the staged output data; and
a memory configured to store the staged output data using the generated masks,
wherein the staged output data comprises a first output data and a second output data, the memory is configured to store merged output data of the first output data and the second output data, and the merged output data includes a merged output segment generated by combining a tail segment from the first output data and a head segment from the second output data when the tail segment and the head segment have complementary masks,
wherein the tail segment is associated with a first mask indicating at least one placeholder value in the tail segment and the head segment is associated with a second mask indicating at least one placeholder value in the head segment, and the merged output segment generated by combining the tail segment and the head segment does not include any placeholder value.
2. The accelerator of claim 1 , wherein the data mask generator comprises a shift register.
3. The accelerator of claim 1 , wherein the pattern generation signal causes the data mask generator to execute a stride based on a type of the output data.
4. The accelerator of claim 1 , wherein the mask has a width equal to an output width of the systolic array.
5. The accelerator of claim 1 , wherein the memory is a dynamic random-access memory.
6. The accelerator of claim 1 , wherein each of the plurality of processing units is a multiply-accumulate unit.
7. The accelerator of claim 1 , wherein the data mask generator is further configured to generate the mask for 2n−1 consecutive writes.
8. The accelerator of claim 1 , wherein the systolic array is further configured to:
receive the merged output data, and
perform a subsequent calculation using the merged output data.
9. The accelerator of claim 1 , wherein the first mask and the second mask have complementary mask patterns.