Using a machine trained network during routing to account for opc cost

Abstract

Some embodiments use a machine-trained network during routing to provide the router with sufficient information to improve the quality of routes generated by a router. This machine-trained network in some embodiments is referred to as the “digital twin” of a lengthy design and/or manufacturing process that produces the design of an IC layout and/or manufactures an IC based on a designed IC layout. The digital twin in some embodiments provides information regarding parasitics, regarding redundant vias for insertion or regarding complexity of subsequent manufacturing processes used to manufacture an IC based on the IC design layout.

1 . For a router that defines a plurality of routes for a plurality of nets in an integrated circuit (IC) design layout, a method comprising:
 performing a first routing operation to define a first set of one or more routes for a first set of one or more nets;   supplying the first set of routes to a machine-trained network (MTN) to identify, for at least one route in the set of route, a set of OPC (Optical Proximity Correction) mask shapes that predicts that OPC mask shapes produced by subsequent OPC operation that is performed to make masks for producing an IC from the design layout that includes the plurality of routes defined by the router;   analyzing the set of OPC mask shapes to produce an OPC cost that quantifies complexity of the subsequent OPC operation;   using the OPC cost to discard one or more routes in the first set of routes; and   performing a second routing operation to define a new route for any net that had a route discarded from the first set of routes.   
     
     
         2 . The method of  claim 1 , wherein the OPC cost quantifies complexity of a set of structures that an OPC operation has to have in the set of masks to produce the IC to have a set of interconnects that implements the first set of routes. 
     
     
         3 . The method of  claim 2 , wherein the set of structures comprises Serifs, hammerheads and line biasing features. 
     
     
         4 . The method of  claim 2 , wherein the OPC operation is an ILT (inverse lithography technology) operation and the set of structures comprises sub-resolution assist features (SRAFs). 
     
     
         5 . The method of  claim 2 , wherein the OPC operation comprises an edge-based OPC operation that produces mask shapes with jagged edges to produce one or more segments of one or more routes, said set of structures comprising one or more shapes with jagged edges. 
     
     
         6 . The method of  claim 1 , wherein using the OPC cost comprises using the OPC cost in a constrained optimization process that searches for an optimal solution for a constrained optimization equation. 
     
     
         7 . The method of  claim 6 , wherein in addition to the OPC cost, the constrained optimization equation accounts for a wirelength cost of the first set of routes. 
     
     
         8 . The method of  claim 7 , wherein the first set of routes consists of a first route for a first net, and the performing the first routing operation, supplying, using and performing the second routing operation are performed to identify a route for the first net. 
     
     
         9 . The method of  claim 7 , wherein the first set of routes comprises at least two routes for at least two nets, and the performing the first routing operation, supplying, using and performing the second routing operation are performed as part of a rip-up-and-reroute operation that considers two or more routes defined for two or more nets to determine whether one or more nets need to have their routes discarded from the first set and new routes defined for them in the second set. 
     
     
         10 . The method of  claim 1 , wherein
 the MTN is trained in a training process that uses a first plurality of known input design layouts with a first plurality of corresponding known output mask layouts comprising the mask shapes used to manufacture components in the corresponding input design layout,   said known input design layouts fed through the MTN during training to produce a second plurality of corresponding output mask layouts comprising predicted mask shapes to manufacture components in the corresponding input design layouts;   the second plurality of output mask layouts used in conjunction with the first plurality of known output mask layouts to generate a loss function value, which is used to adjust a set of trainable parameters of the MTN.   
     
     
         11 . A non-transitory machine readable medium storing a program for execution by at least one processing unit of a router that defines a plurality of routes for a plurality of nets in an integrated circuit (IC) design layout, the program comprising sets of instructions for:
 performing a first routing operation to define a first set of one or more routes for a first set of one or more nets;   supplying the first set of routes to a machine-trained network (MTN) to identify, for at least one route in the set of route, a set of OPC (Optical Proximity Correction) mask shapes that predicts that OPC mask shapes produced by subsequent OPC operation that is performed to make masks for producing an IC from the design layout that includes the plurality of routes defined by the router;   analyzing the set of OPC mask shapes to produce an OPC cost that quantifies complexity of the subsequent OPC operation;   using the OPC cost to discard one or more routes in the first set of routes; and   performing a second routing operation to define a new route for any net that had a route discarded from the first set of routes.   
     
     
         12 . The non-transitory machine readable medium of  claim 11 , wherein the OPC cost quantifies complexity of a set of structures that an OPC operation has to have in the set of masks to produce the IC to have a set of interconnects that implements the first set of routes. 
     
     
         13 . The non-transitory machine readable medium of  claim 12 , wherein the set of structures comprises Serifs, hammerheads and line biasing features. 
     
     
         14 . The non-transitory machine readable medium of  claim 12 , wherein the OPC operation is an ILT (inverse lithography technology) operation and the set of structures comprises sub-resolution assist features (SRAFs). 
     
     
         15 . The non-transitory machine readable medium of  claim 12 , wherein the OPC operation comprises an edge-based OPC operation that produces mask shapes with jagged edges to produce one or more segments of one or more routes, said set of structures comprising one or more shapes with jagged edges. 
     
     
         16 . The non-transitory machine readable medium of  claim 11 , wherein the set of instructions for using the OPC cost comprises a set of instructions for using the OPC cost in a constrained optimization process that searches for an optimal solution for a constrained optimization equation. 
     
     
         17 . The non-transitory machine readable medium of  claim 16 , wherein in addition to the OPC cost, the constrained optimization equation accounts for a wirelength cost of the first set of routes. 
     
     
         18 . The non-transitory machine readable medium of  claim 17 , wherein the first set of routes consists of a first route for a first net, and the sets of instructions for performing the first routing operation, supplying, using and performing the second routing operation are performed to identify a route for the first net. 
     
     
         19 . The non-transitory machine readable medium of  claim 17 , wherein the first set of routes comprises at least two routes for at least two nets, and the sets of instructions for performing the first routing operation, supplying, using and performing the second routing operation are performed as part of a rip-up-and-reroute operation that considers two or more routes defined for two or more nets to determine whether one or more nets need to have their routes discarded from the first set and new routes defined for them in the second set. 
     
     
         20 . The non-transitory machine readable medium of  claim 11 , wherein
 the MTN is trained in a training process that uses a first plurality of known input design layouts with a first plurality of corresponding known output mask layouts comprising the mask shapes used to manufacture components in the corresponding input design layout,   said known input design layouts fed through the MTN during training to produce a second plurality of corresponding output mask layouts comprising predicted mask shapes to manufacture components in the corresponding input design layouts;   the second plurality of output mask layouts used in conjunction with the first plurality of known output mask layouts to generate a loss function value, which is used to adjust a set of trainable parameters of the MTN.