Eliyan has successfully secured $60 million in funding for its innovative chiplet interconnect technology, designed to enhance the speed and efficiency of AI chip processing.
Led by Samsung Catalyst Fund and Tiger Global Management, this funding round aims to tackle the complexities associated with generative AI chip development. With rising demand for AI chips, industry experts predict a staggering 331% growth in the high-bandwidth memory (HBM) sector this year, followed by a 124% increase in 2025, according to Arete Research.
Eliyan’s Universal Chiplet Interconnect Express (UCIe)-, Bandwidth-over-Wire (BoW)-, and Universal Memory Interface (UMI)-compliant physical layer (PHY), known as NuLink PHY, effectively addresses memory and input/output bottlenecks using both advanced and standard packaging materials. This PHY layer connects link layer devices (often termed MACs) to physical media such as optical fibers and copper cables, and is now being utilized in multi-chip solutions that enable the integration of multiple chiplets within a single device.
According to Eliyan, its chiplet interconnect technology offers up to four times the performance and half the power consumption compared to existing solutions. NuLink PHY is validated on advanced process nodes, efficiently handling both die-to-die and die-to-memory interconnects with exceptional performance metrics.
Returning investors, including Intel Capital, SK Hynix, Cleveland Avenue, and Mesh Ventures, also participated in this funding round. This investment follows Eliyan's $40 million Series A round in 2022 and will help advance the company’s focus on the critical challenges in designing and manufacturing next-generation AI chips that utilize multi-die architectures, whether in advanced packaging or standard organic substrates.
In its efforts to tackle the “memory wall” issue in large multi-die designs, Eliyan's innovative UMI delivers a bi-directional interconnect solution that enhances both memory capacity and bandwidth in AI chips.
Marco Chisari, head of Samsung Semiconductor Innovation Center, expressed enthusiasm for the investment: “We are excited to co-lead Eliyan’s Series B round and collaborate with a team renowned for their expertise in interconnect and mixed-signal technologies. The increasing demands of intensive workloads, including generative AI and automotive applications, are driving the need for sophisticated semiconductor designs and the adoption of chiplet architecture.”
The UMI technology significantly optimizes memory bandwidth efficiency in both standard organic substrates and advanced packaging. Its efficient PHY footprint enhances overall memory bandwidth per AI chip while reducing the die area required for memory interfaces.
Srini Ananth, managing director at Intel Capital, noted, “As AI continues to expand connectivity demands and the semiconductor industry shifts toward multi-die implementations, Eliyan is positioned to revolutionize chiplet connectivity. Their advancements in die-to-die interconnect architecture mark a significant milestone in the chiplet revolution and the AI era.”
Eliyan’s NuLink PHY recently taped out on TSMC’s 3nm process, targeting an industry-leading performance of up to 64 Gbps per link, coupled with an unparalleled performance-to-power ratio.
CEO Ramin Farjadrad stated, “This investment underscores confidence in our approach to integrating multi-chip architectures. We are committed to addressing the critical challenges of high costs, low yield, power consumption, manufacturing complexity, and size limitations. Our NuLink technology is commercially ready, optimized for high bandwidth, low latency, and low power capabilities. We deeply appreciate our investors' support in realizing our vision for advanced chiplet systems in the AI era.”
Ranking
