Difference between revisions of "Hybrid Compute-in-Memory: abstract simplified (17695824)"

From WikiPatents
Jump to navigation Jump to search
(Creating a new page)
(Creating a new page)
 
Line 1: Line 1:
−
A compute-in-memory array is a technology that combines computing and memory functions in a single chip. In this case, the array is designed to implement a filter for a specific layer in a neural network.
+
A compute-in-memory array is described in the abstract. This array is designed to implement a filter for a specific layer in a neural network. The purpose of this filter is to process data by multiplying a set of activation bits with a set of filter weight bits. This multiplication process is performed for each channel within a group of channels. The resulting products are then accumulated through a charge accumulation process using multiple capacitors. Finally, the accumulated charge is digitized to produce the output of the filter.
−
 
 
−
The filter operates by multiplying activation bits (binary values indicating the presence or absence of a signal) with filter weight bits (binary values representing the strength of the filter) for each channel in a set of channels. This multiplication process is performed using a charge accumulation method, which involves storing and accumulating electric charge in a group of capacitors.
 
−
 
 
−
Once the charge accumulation is complete, the accumulated charge is converted into digital form (digitized) to produce the output of the filter. This output can then be used in subsequent stages of the neural network for further processing or analysis.
 

Latest revision as of 16:20, 1 October 2023

A compute-in-memory array is described in the abstract. This array is designed to implement a filter for a specific layer in a neural network. The purpose of this filter is to process data by multiplying a set of activation bits with a set of filter weight bits. This multiplication process is performed for each channel within a group of channels. The resulting products are then accumulated through a charge accumulation process using multiple capacitors. Finally, the accumulated charge is digitized to produce the output of the filter.

Retrieved from "http://wikipatents.org/index.php?title=Hybrid_Compute-in-Memory:_abstract_simplified_(17695824)&oldid=416"