- CogniFiber, specializing in photonic computing for AI, raised $5 million led by Chartered Group and Eastern Epic Capital.
- Total funding now stands at $14 million, supporting development of DeepLight processors.
- Founded by Dr. Eyal Cohen and Professor Zeev Zalevsky in 2018, CogniFiber aims to revolutionize data processing with multi-core optical fibers.
- DeepLight technology achieves speeds 1,000 times faster than traditional silicon processors, with energy efficiency as low as 180 watts.
- Plans include launching a photonic supercomputer by 2027 capable of 50 exaflops, significantly reducing AI data center footprint and costs.
Main AI News:
CogniFiber, a trailblazer in photonic computing technology for artificial intelligence (AI) under its DeepLight initiative, proudly announced the successful closure of a $5 million funding round on Thursday. This funding round, prominently led by Chartered Group and Eastern Epic Capital, marks a significant milestone for the Israeli startup, bringing their total raised capital to $14 million, which includes support from the Israel Innovation Authority.
Founded in 2018 by Dr. Eyal Cohen, a distinguished hardware and algorithms engineer with previous affiliations at Mellanox, Saifun, and Oren Semiconductor, and Professor Zeev Zalevsky, an internationally recognized optics researcher and the current Dean of the Faculty of Engineering at Bar-Ilan University, CogniFiber stands at the forefront of photonic computing innovation. The company, currently employing a dedicated team of 13 professionals, is advancing DeepLight, a groundbreaking class of photonic processors designed to revolutionize data processing capabilities.
DeepLight processors harness the power of multi-core optical fibers to process data at the speed of light as it travels through the fiber. This technology enables the execution of AI tasks at unprecedented speeds, surpassing traditional silicon-based systems by up to 1,000 times. Notably, in recent performance tests using their Aurora photonic demonstration system, CogniFiber achieved the remarkable feat of processing 100 million tasks per second with an energy consumption of only 180 watts, setting a new benchmark in efficiency and performance for AI computing.
Looking ahead, CogniFiber aims to deliver a photonic supercomputer capable of achieving up to 50 exaflops by 2027, with a remarkably low power consumption of less than five kilowatts. This advancement holds the promise of significantly reducing the footprint of AI data centers, potentially shrinking installations from the size of a football field to just a few server racks. Such efficiency gains could translate into substantial cost savings, lower electricity consumption by tens of megawatts, and cut down system construction times by over 50%.
The photonic supercomputer developed by CogniFiber offers a compelling alternative or complementary solution to conventional silicon-based computers, which struggle with the growing demands of AI workloads, particularly with the rise of applications like generative AI.
Dr. Eyal Cohen, CEO and Co-Founder of CogniFiber, expressed optimism about the implications of the new funding, stating, “The additional capital will play a crucial role in completing the development of our first-generation processors and scaling our marketing and sales efforts as we prepare to introduce our initial products to global markets. We believe that DeepLight technology will unlock new frontiers in scalable photonic computing, accelerating progress across diverse sectors, from AI-driven smart cities to adaptive cybersecurity and beyond.”
Ana Lipnik-Levy, Vice President of Strategy at CogniFiber, highlighted the challenges associated with introducing a new computing platform to the market but underscored the transformative potential of their technology. “We recognize the importance of market education as we introduce our groundbreaking technology. With the support from this funding round, we are well-positioned to engage strategic partners, penetrate global markets, and pave the way for a new era of AI innovation.”
Conclusion:
CogniFiber’s latest funding round marks a pivotal moment in the advancement of photonic computing for AI applications. With significant strides in speed and energy efficiency, their DeepLight technology not only promises to revolutionize AI data processing but also offers a compelling alternative to traditional silicon-based systems. As they prepare to scale their innovations globally, CogniFiber is poised to reshape the landscape of AI infrastructure, potentially setting new standards for efficiency and performance in the market.
