18063408. REINFORCEMENT LEARNING BASED CORRECTION OF TIMING FAILURES simplified abstract (INTERNATIONAL BUSINESS MACHINES CORPORATION)

From WikiPatents
Jump to navigation Jump to search

REINFORCEMENT LEARNING BASED CORRECTION OF TIMING FAILURES

Organization Name

INTERNATIONAL BUSINESS MACHINES CORPORATION

Inventor(s)

Gregor Boronowsky of Böblingen (DE)

Marvin Von Der Ehe of Vellmar (DE)

Manuel Beck of Dettingen unter Teck (DE)

Jan Niklas Stegmaier of Sindelfingen (DE)

Simon Hermann Friedmann of Böblingen (DE)

REINFORCEMENT LEARNING BASED CORRECTION OF TIMING FAILURES - A simplified explanation of the abstract

This abstract first appeared for US patent application 18063408 titled 'REINFORCEMENT LEARNING BASED CORRECTION OF TIMING FAILURES

Simplified Explanation

The abstract describes a computer implemented method of correcting timing failures in an integrated circuit design using a reinforcement learning agent with a neural network.

  • The method involves receiving a graph encoding the network of conductors and repowering structures, and receiving modification recommendations from the reinforcement learning agent based on the input graph.
    • Potential Applications:**

- Semiconductor industry for improving integrated circuit designs - Electronics manufacturing for enhancing circuit performance and reliability

    • Problems Solved:**

- Correcting timing failures in network of conductors - Optimizing integrated circuit design for better performance

    • Benefits:**

- Improved timing and performance of integrated circuits - Automated correction of timing failures using reinforcement learning

    • Potential Commercial Applications:**

- Semiconductor companies for optimizing circuit designs - Electronics manufacturers for enhancing product performance

    • Possible Prior Art:**

- Prior art related to reinforcement learning in integrated circuit design optimization - Previous methods for correcting timing failures in network of conductors

    • Unanswered Questions:**
    • 1. How does the neural network in the reinforcement learning agent learn to recommend modifications for correcting timing failures in the integrated circuit design?**

- The abstract does not provide details on the specific training process or algorithms used by the neural network to make modification recommendations.

    • 2. Are there any limitations or constraints in the application of this method to different types of integrated circuit designs or network structures?**

- The abstract does not mention any potential limitations or constraints that may arise when applying this method to diverse integrated circuit designs or network configurations.


Original Abstract Submitted

Disclosed herein is a computer implemented method of correcting a timing failure of a network of conductors and repowering structures in an integrated circuit design using a reinforcement learning agent. The reinforcement learning agent comprises a neural network. The method comprises: receiving a graph comprising nodes and edges that encodes said network of conductors and repowering structures; and receiving a modification recommendation from said reinforcement learning agent in response to inputting said graph into said reinforcement learning agent.

Retrieved from "http://wikipatents.org/index.php?title=18063408._REINFORCEMENT_LEARNING_BASED_CORRECTION_OF_TIMING_FAILURES_simplified_abstract_(INTERNATIONAL_BUSINESS_MACHINES_CORPORATION)&oldid=58748"