Sparse codec for neural network
AssigneeMoffett International Co., Limited
InventorChangxu ZHANG
Life of this patent
Estimating the term…
Loading patent…
AssigneeMoffett International Co., Limited
InventorChangxu ZHANG
Estimating the term…
An advanced processor is disclosed for efficient handling of sparse tensor data. The processor comprises an input buffer configured to receive an input tensor, a gather circuit that collects high-magnitude tensor data from a dense tensor to form a condensed tensor, and a scatter circuit that distributes tensor data from a condensed tensor into a sparse uncondensed tensor based on a given mask. The processor is configured to perform sparse encoding, generating a sparse representation of the input tensor that includes: (1) the condensed tensor containing high-magnitude tensor data gathered by the gather circuit, and (2) a sparse mask indicating the position information of high-magnitude tensor data in the input tensor. Additionally, the processor performs sparse decoding by scattering the condensed tensor into the sparse uncondensed tensor using the scatter circuit and the sparse mask.
A processor, comprising:
A method for sparse encoding a tensor, comprising:
A method for sparse decoding a sparse representation of a tensor, comprising:
Dependencies are inferred from the claim text.