LightSolver debuts ‘laser processing unit’ as alternative to quantum … – SiliconANGLE News

High-performance computing startup LightSolver Ltd. today announced what it says is the worlds first pure laser-based processing unit, providing businesses with an alternative to expensive supercomputers and unstable quantum computing platforms.

The LightSolver LPU is designed to solve business challenges that involve multiple variables. Defined as NP-hard optimization problems, they traditionally require specific mathematical models and vast amounts of computational power to solve.

LightSolver said these optimization problems are commonly faced in industries such as finance, centered around portfolio management, risk management and trading optimization. Its platform also targets problems around resource allocation, assembly line balancing and facility layout optimization in the manufacturing industry, and engineering challenges such as complex material design.

Organizations have traditionally relied on supercomputers to solve these challenges, and are also invested in quantum computers, which have the potential to become vastly more powerful. However, LightSolver says supercomputers are close to approaching their computational limits, while quantum computers are neither scalable or practical yet. Both options are incredibly expensive and require significant expertise and resources, it says.

LightSolver is different, because its LPU platform is accessible as a software-as-a-service for any organization. The startup said its quantum-inspired technology employs coupled lasers to perform extremely complex computations for intricate multi-variable problems. Built from well-understood laser technology, it can provide solutions to business optimization problems that were previously considered to be unsolvable, the company said.

Co-founder and Chief Executive Ruti Ben-Shlomi (pictured, right, alongside co-founderChene Tradonsky) told SiliconANGLE that LightSolver is being made available via a user-friendly interface through a Python package.

Many large and medium-sized companies are already familiar with mapping their business challenges as mathematical optimization problems, she explained. For organizations that do not have these capabilities yet, we offer assistance in translating their business challenges into mathematical formulations that can be solved by our SaaS.

The LPU technology is said to mimic the power of quantum computing, using all-optical lasers that require no electronics to compute. Moreover, its platform is available in an extremely small form factor, no bigger than a desktop computer, meaning it has low power requirements and operates at room temperature.

Just as graphics processing units have surpassed the capabilities of central processing units, LPUs can outperform the most powerful supercomputers, LightSolver claims. They work at the speed of light to calculate every possible variable in a problem instantaneously. For any given problem, the LPU will convert a mathematical representation into a physical logic formulation, then map it into obstacles within the lasers optical path. The laser beams will then converge into the most practical solution, which is then translated back into the business language of the user.

LightSolver says its great advantage is its ability to harness the natural properties of light to break the physical limits of electronics, enabling it to solve problems faster and with more accuracy than any supercomputer.

Ben-Shlomi was keen to point out some of the specific use cases where LightSolver can deliver a significant advantage to companies. For instance, he said that in industries such as aviation and aerospace, material manufacturing involves finding the optimized composition of materials to meet multiple constraints, such asrigidity, heat capacity and resistance.

This process can be challenging due to the complex nature of these constraints, Ben-Shlomi said. LightSolver provides fast and accurate solutions for these problems, with the ability to scale to a large number of variables.

It can also help with tasks such as warehouse optimization. If a customer has a representative mathematical model for the variables affecting the layout of its warehouses, it can simply input this into LightSolver to find the most streamlined option. If you dont have such a model yet, we can work with you to develop one that fits your specific needs, Ben-Shlomi added. Once the model is developed, you can then run it on our SaaS.

LightSolvers claims of superiority over supercomputers are backed up by a number of academic research papers. For instance, in a head-to-head challenge against an industry-leading deep learning solver, its platform was able to solve Max-2-SAT problems by up to 1,000 times faster. A second paper reveals how LightSolver was used to develop a quantum-inspired algorithm for sparse coding that resulted in more accurate estimations than classical approximation methods.

Ben-Shlomi said LightSolver will be able to maintain its advantage even when quantum computers become more mature, with the ability to scale up to thousands of variables. We can provideour solutions using a small-sized rack unit running at room temperature with low power consumption, she said. This means that we can leverage the power of quantum computing without the need for large, expensive hardware or specialized cooling systems and solve complex optimization problems more efficiently and cost-effectively.

TheCUBEis an important partner to the industry. You guys really are a part of our events and we really appreciate you coming and I know people appreciate thecontent you create as well Andy Jassy

THANK YOU

View original post here:
LightSolver debuts 'laser processing unit' as alternative to quantum ... - SiliconANGLE News

Related Posts

Comments are closed.