[ad_1]
Lightmatter Moves to Take on the AI Computation Market
Lightmatter, a photonic computing start-up, is entering the rapidly growing AI computation market with a hardware-software combo it claims will help the industry level up — and save a lot of electricity to boot. Lightmatter’s chips use optical flow to solve computational processes like matrix vector products. Traditionally, this math has been performed by GPUs and TPUs that specialise in it but use traditional silicon gates and transistors. The issue with those is that we’re approaching the limits of density and therefore speed for a given wattage or size.
Advances are still being made but at great cost and pushing the edges of classical physics.
Lightmatter aims to be a leader in new compute paradigms. Its approach is faster and more efficient, using arrays of microscopic optical waveguides to let light essentially perform logic operations just by passing through them. Since the waveguides are passive, the main power draw is creating the light itself, then reading and handling the output.
The Challenges of the AI Scalability Market
The world’s biggest companies are hitting an energy power wall and experiencing massive challenges with AI scalability. Traditional chips push the boundaries of what’s possible to cool, and data centres produce increasingly large energy footprints. Some have projected that training a single large language model can take more energy than 100 U.S. homes consume in a year. Additionally, it’s estimated that 10% to 20% of the world’s total power will go to AI inference by the end of the decade unless new compute paradigms are created.
Lightmatter’s Solutions
Lightmatter’s solution is potentially a game-changer for the industry. The optical computing process increases the chip’s power just by using more than one color at once, or in other words, two colors will double the power. The company has a full stack of computing hardware (Envise), interconnect (Passage) and software (Idiom) — the latter of which is said to let machine learning developers adapt quickly.
Seamless Integration for Developers
Lightmatter has built a software stack that integrates seamlessly with PyTorch and TensorFlow. The workflow for machine learning developers remains the same: the neural networks built with these industry standard applications import their libraries, allowing all the code to run on Envise.
Pilots and Mass Production
Lightmatter has several pilots running, and mass production is planned for 2024. The funding for this round came from SIP Global, Fidelity Management & Research Company, Viking Global Investors, GV, HPE Pathfinder and existing investors.
FAQs
What is Lightmatter?
Lightmatter is a photonic computing start-up that is aiming to change the game in AI computation.
What is its solution to the challenges in the AI scalability market?
Lightmatter’s solution is hardware and software built to work together. Its chips use arrays of microscopic optical waveguides to let light perform logic operations just by passing through them, making it faster and more efficient than traditional silicon gates and transistors.
When is mass production planned for?
Mass production is scheduled for 2024.
Who are Lightmatter’s investors?
SIP Global, Fidelity Management & Research Company, Viking Global Investors, GV, HPE Pathfinder, and existing investors have all invested in Lightmatter.
[ad_2]
For more information, please refer this link