“Cutting-Edge Chip Breakthrough Promises Enhanced AI Efficiency!”
By Stephen NellisSAN FRANCISCO (Reuters) – A Silicon Valley startup called Lightmatter has unveiled a revolutionary computer chip that can accelerate artificial intelligence tasks while consuming less energy. The company, now valued at $4.4 billion following an $850 million venture capital raise, is at the forefront of leveraging light beams instead of electronic signals to facilitate faster data transfer between computers. This high-speed connectivity is crucial for AI applications that require extensive computational resources spread across multiple machines.Lightmatter’s innovative approach is detailed in a recent paper published in the prestigious scientific journal Nature. Unlike traditional computers that rely on transistors, Lightmatter’s chip utilizes meticulously controlled light beams that interact with each other and are analyzed by integrated chips produced by its manufacturing partner, GlobalFoundries. This novel method avoids the challenges associated with shrinking transistors in conventional chips, offering a promising alternative for future computing advancements.Previous photonic computing endeavors have struggled with precision, often leading to inaccurate outcomes. Lightmatter addresses this issue by partitioning large and small numbers before processing them through its photonic circuits, ensuring that minute values are preserved. Nick Harris, the CEO of Lightmatter, shared that the chip currently matches the precision of existing AI processors and anticipates that it will take around a decade for this technology to become mainstream.”Our focus is on the evolution of processors and the endless possibilities they hold. This unconventional approach represents a significant opportunity in the advancement of computing technology, with substantial economic value driving the continuous improvement of computers,” remarked Harris.(Reporting by Stephen Nellis in San Francisco; Editing by Stephen Coates)
