- Lightolver unveils laser processor unit designed to resolve physical equations at high speed
- The laser computer system uses optical memory to avoid processor and memory bottlenecks
- The company is starting to offer access to the early laboratory with a roadmap to a million variables by 2029
Laser -based IT is becoming a potential challenger to establish high performance computer systems, with a light of lights revealing what it calls a laser processing unit (LPU).
The startup based in Tel Aviv indicates that its equipment can directly resolve partial differential equations, a fundamental class of problems in physics and engineering.
Unlike quantum systems or GPU clusters, the approach is based on a lasers grid acting in unison.
Cartography of equations on the LPU
Lightolver says that its material can map equations such as the thermal equation (which models the heat flow) and the Schrödinger equation (a basic equation in quantum mechanics) directly on the LPU.
Using the natural properties of lasers as electromagnetic waves, Lightolver technology avoids many constraints found in digital systems.
This includes the bottlenecks caused by data transfer between memory and the processor.
According to Lightolver, the LPU works with an integrated optical memory. In practice, this means that laser states are kept inside the resonator, allowing each calculation stage to rely on the last without having to move the data outside.
The result is a constant time iteration measured in nanoseconds, regardless of the size of the problem.
While Lightolver compares its platform with high-performance and quantum computers, he emphasizes that it is not built on photonic fleas.
Today’s photonic processors are two -dimensional, while its laser -based design is three -dimensional, which, according to the startup, will allow greater scalability.
Lightolver has established a roadmap to reach 100,000 variables by 2027 and one million by 2029.
The company is already starting to make its equipment available for tests. An alpha version of its processor and a digital emulator are offered to researchers through what he calls his LPU laboratory.
Lightolver believes that this step will help scientists and engineers to experiment with technology before reaching the trade note.
“Classic computers scan analog nature, and we pay the price in longer hours and waste energy. By performing large-scale physics simulations on a physical machine, we can resolve them more efficiently than any HPC or quantum system available today, ”said Ruti Ben-Shlomi, CEO and Co-founder of Lightolver.
Research articles were published on the subject and presentations were presented during computer conferences, in particular ACM Computing Frontiers 2025.
The company has also created partnerships with suppliers of simulation software and works with HPC centers and national laboratories.
