System and method for determining hybrid-manufacturing process plans for integrated circuits based on satisfiability modulo difference logic solver

Patent metadata

Field	Value
Publication number	US-11157672-B1
Application number	US-202017023747-A
Country	US
Kind code	B1
Filing date	Sep 17, 2020
Priority date	Sep 17, 2020
Publication date	Oct 26, 2021
Grant date	Oct 26, 2021

Abstract

One embodiment of the present disclosure provides a system for determining a hybrid-manufacturing plan for manufacturing an integrated circuit (IC). During operation, the system can obtain a set of hybrid-manufacturing constraints for manufacturing the IC. The set of hybrid-manufacturing constraints can include a set of primitives, a set of atoms, and an atom end-state vector. An atom can correspond to a unit of spatial volume of the IC. A primitive can represent an additive, subtractive, or a mixed manufacturing process corresponding to one or more atoms of the IC. Next, the system can determine a plurality of feasible hybrid-manufacturing plans based on the set of manufacturing constraints. Each feasible hybrid-manufacturing plan can represent an ordering of the set of primitives that satisfies the atom end-state vector. The system can then determine costs for manufacturing the IC using the plurality feasible hybrid-manufacturing plans. The system can determine, based on the costs, an optimized hybrid-manufacturing plan for manufacturing the IC.

First claim

What is claimed is:

1. A method for determining a hybrid-manufacturing plan for an integrated circuit (IC), the method comprising:
obtaining, by a computer, a set of hybrid-manufacturing constraints for manufacturing the IC, wherein the set of hybrid-manufacturing constraints include a set of primitives, a set of atoms, and an atom end-state vector, wherein an atom corresponds to a unit of spatial volume of the IC, and wherein a primitive represents an additive, subtractive, or a mixed manufacturing process corresponding to one or more atoms of the IC;
determining a plurality of feasible hybrid-manufacturing plans based on the set of hybrid-manufacturing constraints, wherein each feasible hybrid-manufacturing plan represents an ordering of the set of primitives that satisfies the atom end-state vector;
determining costs for manufacturing the IC using the plurality of feasible hybrid-manufacturing plans; and
determining, based on the costs, an optimized hybrid-manufacturing plan for manufacturing the IC.

2. The method of claim 1 , wherein the set of hybrid manufacturing constraints further comprises:
a constraint matrix with the columns corresponding to the set of primitives and the rows corresponding to the set of atoms; and
an atom cost vector.

3. The method of claim 1 , further comprising converting the set of hybrid-manufacturing constraints to a satisfiability modulo theory (SMT) problem;
wherein determining the feasible hybrid-manufacturing plans comprises solving the SMT problem using a Satisfiability (SAT) modulo difference logic.

4. The method of claim 1 , wherein each cost is associated with one or more of:
tool set-up cost;
shut-down cost; and
material cost.

5. The method of claim 1 , wherein the primitive is further categorized into:
a preparatory manufacturing process; and
a post-processing manufacturing process; and
wherein each manufacturing process includes one or more of:
a lithography exposure process;
a chemical bath process; and
a plasma etching process.

6. The method of claim 1 , further comprising converting the set of hybrid-manufacturing constraints to an SMT problem by:
generating a conjunctive normal form (CNF) modulo difference logic formula; and
converting the CNF difference logic formula to a Boolean formula, wherein the Boolean formula is implemented by using at least one or more Boolean subtractor circuits.

7. The method of claim 6 , wherein determining the plurality of feasible hybrid-manufacturing plans further comprises:
for a respective feasible hybrid-manufacturing plan:
sorting one or more columns in a constraint matrix based on one or more variables in the feasible hybrid-manufacturing plan, wherein each column corresponds to a primitive, and wherein sorting the one or more columns in the constraint matrix corresponds to changing an order of the set of primitives;
determining, for each primitive, an atom count representing a number of atoms added, deleted, or modified;
determining an atom cost associated with each primitive; and
determining a cost of the feasible hybrid-manufacturing plan by multiplying the atom cost and the atom count for respective primitives and aggregating across the set of primitives.

8. The method of claim 1 , wherein determining the optimized hybrid-manufacturing plan further comprises:
applying a binary search to the costs to determine the optimized hybrid-manufacturing plan.

9. A non-transitory computer-readable storage medium storing instructions that when executed by a computer cause the computer to perform a method for determining a hybrid-manufacturing plan for manufacturing an integrated circuit (IC), the method comprising:
obtaining a set of hybrid-manufacturing constraints for manufacturing the IC, wherein the set of hybrid-manufacturing constraints include a set of primitives, a set of atoms, and an atom end-state vector, wherein an atom corresponds to a unit of spatial volume of the IC, and wherein a primitive represents an additive, a subtractive, or a mixed manufacturing process corresponding to one or more atoms of the IC;
determining a plurality of feasible hybrid-manufacturing plans based on the set of hybrid-manufacturing constraints, wherein each feasible hybrid-manufacturing plan represents an ordering of the set of primitives that satisfies the atom end-state vector;
determining costs for manufacturing the IC using the plurality of feasible hybrid-manufacturing plans; and
determining, based on the costs, an optimized hybrid-manufacturing plan for manufacturing the IC.

10. The non-transitory computer-readable storage medium of claim 9 , wherein the set of hybrid manufacturing constraints further comprises:
a constraint matrix with the columns corresponding to the set of primitives and the rows corresponding to the set of atoms; and
an atom cost vector.

11. The non-transitory computer-readable storage medium of claim 9 , wherein the primitive is categorized into:
a preparatory manufacturing process; and
a post-processing manufacturing process; and

wherein each manufacturing process includes one or more of:
a lithography exposure process;
a chemical bath process; and
a plasma etching process.

12. The non-transitory computer-readable storage medium of claim 9 , wherein the method further comprises converting the set of hybrid-manufacturing constraints to a satisfiability modulo theory (SMT) problem; and
wherein determining the feasible hybrid-manufacturing plans comprises solving the SMT problem using a Satisfiability (SAT) modulo difference logic.

13. The non-transitory computer-readable storage medium of claim 9 , wherein the method further comprises converting the set of hybrid-manufacturing constraints to an SMT problem by:
generating a conjunctive normal form (CNF) modulo difference logic formula; and
converting the CNF difference logic formula to a Boolean formula, wherein the Boolean formula is implemented by using at least one or more Boolean subtractor circuits.

14. The non-transitory computer-readable storage medium of claim 13 , wherein determining the plurality of feasible hybrid-manufacturing plans further comprises:
for a respective feasible hybrid-manufacturing plan:
sorting one or more columns in a constraint matrix based on one or more variables in the feasible hybrid-manufacturing plan, wherein each column corresponds to a primitive, and wherein sorting the one or more columns in the constraint matrix corresponds to changing an order of the set of primitives;
determining, for each primitive, an atom count representing a number of atoms added, deleted, or modified;
determining an atom cost associated with each primitive; and
determining a cost of the feasible hybrid-manufacturing plan by multiplying the atom cost and the atom count for respective primitives and aggregating across the set of primitives.

15. The non-transitory computer-readable storage medium of claim 9 , wherein determining the optimized hybrid-manufacturing plan further comprises:
applying a binary search to the costs to determine the optimized hybrid-manufacturing plan.

16. A computer system for determining a hybrid-manufacturing plan for manufacturing an integrated circuit, the system comprising:
a processor; and
a storage device coupled to the processor and storing instructions, which when executed by the processor cause the processor to perform a method, wherein the method comprises:
obtaining a set of hybrid-manufacturing constraints for manufacturing the IC, wherein the set of hybrid-manufacturing constraints include a set of primitives, a set of

Assignees

• Palo Alto Res Ct Inc

Inventors

• Feldman Aleksandar B
• Behandish Morad
• De Kleer Johan
• Matei Ion
• Nelaturi Saigopal

Classifications

• G06F2111/04

  Constraint-based CAD · CPC title

• G06F30/39

  Circuit design at the physical level (physical level design for reconfigurable circuits G06F30/347) · CPC title

• G06F2119/18

  Manufacturability analysis or optimisation for manufacturability · CPC title

• G05B2219/45031

  Manufacturing semiconductor wafers · CPC title

• G05B19/41865 Primary

  characterised by job scheduling, process planning, material flow · CPC title