17643256. SOLVING OPTIMIZATION PROBLEMS WITH PHOTONIC CROSSBARS simplified abstract (INTERNATIONAL BUSINESS MACHINES CORPORATION)

SOLVING OPTIMIZATION PROBLEMS WITH PHOTONIC CROSSBARS

Organization Name

INTERNATIONAL BUSINESS MACHINES CORPORATION

Inventor(s)

Ghazi Sarwat Syed of Zurich (CH)

Abu Sebastian of Adliswil (CH)

SOLVING OPTIMIZATION PROBLEMS WITH PHOTONIC CROSSBARS - A simplified explanation of the abstract

This abstract first appeared for US patent application 17643256 titled 'SOLVING OPTIMIZATION PROBLEMS WITH PHOTONIC CROSSBARS

Simplified Explanation

The invention is a method for optimizing a photonic crossbar array structure that includes N input lines and M output lines interconnected by photonic memory devices. The memory devices are programmed to store weights for solving an optimization problem. The method operates as follows:

• Determine values of L input vectors of N components each.
• Generate N electromagnetic signals based on the determined values.
• Each generated signal multiplexes L input signals encoded at respective wavelengths.
• Apply the N electromagnetic signals to the N input lines of the photonic crossbar array structure.

Potential applications of this technology include:

• Optimization problems in various fields such as logistics, finance, and telecommunications.
• Data processing and analysis in large-scale systems.
• Machine learning and artificial intelligence algorithms.

Problems solved by this technology:

• Solving complex optimization problems efficiently.
• Handling large amounts of data and processing it in parallel.
• Improving the speed and accuracy of optimization algorithms.

Benefits of this technology:

• Faster and more efficient optimization solutions.
• Scalability to handle large-scale problems.
• Parallel processing capabilities for improved performance.
• Potential for integration with other technologies such as machine learning and AI.

Original Abstract Submitted

The invention is directed to solving an optimization problem. The method operates a photonic crossbar array structure including N input lines and M output lines, which are interconnected at junctions via N×M photonic memory devices, where N≥2 and M≥2. The photonic memory devices are programmed to store respective weights in accordance with the optimization problem. The photonic crossbar array structure is operated as follows. First, the method determines values of L input vectors of N components each, where L≥2. Second, based on the determined values, N electromagnetic signals are generated, where each of the generated signals multiplexes L input signals encoded at respective wavelengths, so as for the N electromagnetic signals to map the L input vectors of N components each. Third, the N electromagnetic signals generated are applied to the N input lines of the photonic crossbar array structure.