Lightmatter
Lightmatter, headquartered in Mountain View, CA, with an office in Boston, MA, is valued at $1.2 billion and offers photonic technology products like Passage, Envise, and Idiom to reduce AI's environmental impact.
Company Overview
Lightmatter, headquartered in Mountain View, CA, with an additional office in Boston, MA, is making significant strides in reducing the environmental impact of artificial intelligence through its photonic technology. As of 2023, the company has raised $310 million in funding, reaching a valuation of $1.2 billion. Lightmatter focuses on innovations in photonic computing, claiming to enable the next Moore's Law with its cutting-edge technology.
Products
Lightmatter offers an array of products designed to enhance AI technology while mitigating its environmental footprint. Their key products include Envise, Passage, and Idiom. Envise is a general-purpose machine learning accelerator that integrates photonics with traditional transistor-based systems. Passage serves as a wafer-scale, programmable photonic interconnect that allows arrays of diverse chips to communicate. Idiom interfaces seamlessly with standard deep learning frameworks, delivering necessary tools for deep learning model authors and deployers.
Envise Machine Learning Accelerator
Envise, one of Lightmatter's flagship products, is a general-purpose machine learning accelerator that uniquely combines photonics and transistor-based systems. This hybrid approach aims to boost performance and efficiency in machine learning applications, contributing to Lightmatter's goal of reducing the environmental impact of AI.
Passage Photonic Interconnect
Passage is another innovative product from Lightmatter, acting as a wafer-scale, programmable photonic interconnect. This technology enables seamless communication between arrays of heterogeneous chips, facilitating advanced computing capabilities and enhancing the efficiency and performance of AI systems.
Idiom Deep Learning Interface
Idiom from Lightmatter is designed to interface with standard deep learning frameworks. It provides the necessary tools required by deep learning model authors and deployers, ensuring that the deployment and development processes are streamlined and efficient.