Optimized hardware implementations for monotonic, non-decreasing thresholding
AssigneeXilinx, Inc.
InventorsYaman UMUROGLU, Nicholas FRASER, Thomas Bernd PREUSSER
Life of this patent
Estimating the term…
Loading patent…
AssigneeXilinx, Inc.
InventorsYaman UMUROGLU, Nicholas FRASER, Thomas Bernd PREUSSER
Estimating the term…
Embodiments herein describe generating non-linear thresholds which can be used, for example, in a quantization operation of a neural network layer. In one example, a compiler can receive a set of non-linear thresholds (e.g., rounded integer thresholds) and determine delta values indicating respective differences between two thresholds of the non-linear thresholds. Because the thresholds are non-linear, the delta values are also different. The compiler can calculate residual errors to represent the differences between the deltas with respect to a step size. Using these residual errors, an initial threshold value, a step size between the thresholds, the non-linear thresholds can be calculated as needed, instead of having to store the non-linear thresholds in memory.
A system comprising:
A non-transitory computer readable medium storing instructions that, when executed by a combination of one or more processors, cause the combination of one or more processors to perform an operation comprising:
A computing system comprising:
Dependencies are inferred from the claim text.