Microsoft Develops Analog Optical Computer for Solving Optimization Problems
Microsoft has unveiled its newly-developed Analog Iterative Machine (AIM), an analog optical computer designed to solve difficult optimization problems. The AIM machine, described as a computer that uses photons and electrons instead of transistors to process data, has the potential to surpass state-of-the-art digital technology and even quantum computers, according to TechSpot.
The AIM machine works by allowing light to pass through several layers, creating impressions on a modular array that replace the function of a standard transistor. This enables the machine to solve practical problems at the speed of light, providing a faster and more efficient solution compared to traditional digital computers.
AIM is built using low-cost and scalable commodity opto-electronic technologies and is based on an asynchronous data flow architecture. It is specifically designed for solving optimization problems, such as the traveling salesman riddle, which are prevalent in industries like finance, logistics, transportation, energy, healthcare, and manufacturing.
Traditional digital computers struggle to solve optimization problems in a timely, energy-efficient, and cost-effective manner due to the exponential growth of possible combinations. AIM aims to address this challenge by offering a 100x increase in performance compared to current digital approaches.
The AIM machine is currently a research project with limited access for potential customers. However, UK financial company Barclays is already testing the machine to track transactions of money into stock purchases. Microsoft has also released an AIM simulator as a service, allowing selected users, including collaborators from Princeton University and Cambridge University, to gain first-hand experience with the technology.
Microsoft’s AIM machine presents a promising alternative to traditional digital computing, offering a new level of speed and efficiency in solving optimization problems. With its ability to process continuous value data, rather than being limited by binary systems, AIM could revolutionize industries that heavily rely on complex mathematical computations.