Laser Focus World: Can you introduce us to LightSolver’s LPU100?
Ruti Ben-Shlomi: The LPU100 is based on our laser processing paradigm, which leverages the unique properties of light to provide high-performance computing power. It unleashes the power of 100 lasers to solve NP-hard problems, challenging the processing times of quantum and supercomputers. Its laser array represents 100 continuous variables and can tackle optimization problems with up to 120100 combinations (see video).
Problems like the classic Travelling Salesman Problem (TSP) can be found nearly everywhere across business and science—from routing and dispatching optimization for supply chain and logistics operations, to risk modeling in finance, and material science in aerospace. But they can’t be solved efficiently and accurately with classical computers or today’s quantum computers.
Our LPU100 is now available to a select number of enterprises through the public cloud.
LFW: Optics/photonics involved?
Ben-Shlomi: LPU100 is a purely optical system and is not photonic computing because we do not use individual photons to carry information. LightSolver uses a laser array in which a mathematical problem is encoded into the phases of the laser. The coupled lasers then interact and seamlessly converge toward a minimal energy-loss state that provides the solution to the problem.
LFW: How is it quantum-inspired?
Ben-Shlomi: One of the main advantages of our laser-based computing method is that it scales and computes much faster than classical processors and can tackle problems that are unfeasible for graphics processing units (GPUs), due to their size and complexity.
To get a sense of this acceleration: LPU100 can execute a vector matrix multiplication—a standard operation in optimization and other applications—in as little as 10 nanoseconds, while a GPU requires multiple microseconds to complete the same operation. This accelerated computation enables organizations to save compute time, which translates into reduced energy costs and carbon footprint.
LPU100 also has several advantages over quantum computers. It is much less demanding in terms of environmental conditions: no vacuum or ultracold temperatures are required. And an LPU is only about the size of a desktop computer, so it can reside in an average server room, aboard vehicles, or on remote sites such as oil rigs, with low operating costs. Most importantly, laser processing is ready to solve real-world problems today!
LFW: Anything surprising about LPU100’s development or functionality?
Ben-Shlomi: One of the most remarkable aspects of our LPU100 is that it harnesses a naturally occurring phenomenon, the interaction of light waves within laser cavities, to execute highly complex computations. Unlike most optical computing methods that process one operation at a time, our approach allows for the execution of an entire algorithm simultaneously.
With the immense funding and extensive research spent over the past few decades exploring quantum and other computing paradigms beyond silicon, it’s fascinating nobody ever considered using lasers to compute. We ventured into uncharted waters, were able to uncover new physics, and ultimately transformed our discovery into a commercial computing platform.
LFW: Applications?
Ben-Shlomi: The LPU100 addresses NP-hard problems, which are found almost everywhere across science and business. LightSolver currently focuses on:
- Logistics to optimize day-to-day operations by improving vehicle routing, scheduling and dispatching, warehouse optimization, and supply chain management in real time.
- Manufacturing to increase productivity in factories and warehouses through optimized resource allocation, scheduling, assembly line balancing, and facility layout optimization.
- Aerospace to optimize material design and enhance decision intelligence for mission-critical processes in which ultrafast processing times are needed.
- Finance to boost returns and decrease risk in critical financial challenges, including portfolio management, trading optimization, liquidity and risk management, and financial planning and budgeting.
LFW: What’s next?
Ben-Shlomi: We’ll continue to scale out our solution to tackle optimization problems of even greater size and complexity.