Multiply-accumulate (MAC) unit for single-instruction/multiple-data (SIMD) instructions
AssigneeIntel Corporation
InventorsDeli Deng, Anthony Jebson, Yuyun Liao, Nigel C. Paver, Steve J. Strazdus
Life of this patent
Estimating the term…
Abstract
A tightly coupled dual 16-bit multiply-accumulate (MAC) unit for performing single-instruction/multiple-data (SIMD) operations may forward an intermediate result to another operation in a pipeline to resolve an accumulating dependency penalty. The MAC unit may also be used to perform 32-bit×32-bit operations.
Claims9
An article comprising a machine-readable medium which stores machine-executable instructions, the instructions causing a machine to:
- perform a first compression operation in a first multiply-accumulate operation in a pipeline;
- generate two or more intermediate vectors in a first compression operation in the first multiply-accumulate operation;
- and
- forward at least a portion of each of the two or more intermediate vectors to a second multiply-accumulate operation in the pipeline prior to completion of the first multiply-accumulate operation.
An article comprising a machine-readable medium which stores machine-executable instructions, the instructions causing a machine to:
- compress a first plurality of partial products into a first sum vector and a first carry vector and compressing a second plurality of partial products into a second sum vector and a second carry vector in a first Wallace tree compression stage of a first multiply-accumulate operation;
- compress the first and second sum vectors and the first and second carry vectors into a first intermediate sum vector and a first intermediate carry vector;
- compress the intermediate sum vector and a third plurality of partial products and compressing the intermediate carry vector and a fourth plurality of partial products in a second stage of the first multiply-accumulate operation;
- and
- forward the intermediate sum and carry vectors to a second multiply-accumulate operation in a pipeline prior to completion of the first multiply-accumulate operation.
Dependencies are inferred from the claim text.